Skip to main content

Full text of "Journal of the Washington Academy of Sciences"

See other formats





VOLUME 38, 1948 

James I. Hoffman Alan Stone Frank C. Kracek 



Lawrence A. Wood Richard E. Blackwelder 


J. P. E. Morrison James S. Williams 


Elbert L. Little, Jr. Waldo R. Wedel 


Irl C. Schoonover 





450 Ahnaip St. 

at Men.asha, Wisconsin 


No. 1, pp. 1-32, January 15, 1948. 
No. 2, pp. 33-80, February 16, 1948. 
No. 3, pp. 81-112, March 13, 1948. 
No. 4, pp. 113-144, April 14, 1948. 
No. 5, pp. 145-192, May 20, 1948. 
No. 6, pp. 193-224, June 21, 1948. 
No. 7, pp. 225-256, August 2, 1948. 
No. 8, pp. 257-288, August 23, 1948. 
No. 9, pp. 289-320, October 1, 1948. 1 
No. 10, pp. 321-352, November 2, 1948. 
No. 11, pp. 353-388, November 12, 1948. 
No. 12, pp. 389-425, December 30, 1948. 

b. 15 

1 Vol.38 

January 15, 1948 

No. 1 






William N. Fenton 


James I. Hoffman 


Alan Stone 



Frank C. Kracek 


Ira B. Hansen 


John A. Stevenson 


Richard E. Blackweldbr 


Ralph W. Imlat 


Waldo R. Wedel 


Irl C. Schoonover 




450 Ahnaip St. 

at Menasha, Wisconsin 

Entered as second class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a special rate of postage provided for in the Act of February 28, 1926, 

Authorised January 21. 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author 1 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate) : 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 .50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10 .00 $0 .70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8 .00 .60 

Vol. 23 to current vol. (12 numbers per vol.) 7 .50 .90 

• Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25 .00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 

Officers of the Academy 

President: Waldo L. Schmitt, U. S. National Museum. 

Secretary: C. Lewis Gazin, U. S. National Museum. 

Treasurer: Howard S. Rappleye, U. S. Coast and Geodetic Survey. 

Archivist: Nathan R. Smith, Bureau of Plant Industry, Soils, and Agricultural 

Custodian of Publications: Harald A. Rehder, U. S. National Museum. 




Vol. 38 

January 15, 1948 

No. 1 

MATHEMATICS.- — The role of the concept of equivalence in the study of physical 
and mathematical systems. 1 Richard Stevens Burington, Bureau of Ord- 
nance, Navy Department. 

Foreword. — Many mathematicians as 
well as engineers and physicists sometimes 
fail to recognize the great principles un- 
derlying mathematical reasoning and meth- 
ods that permeate their particular fields. 
Instead, they are apt to turn their attention 
solely to the formalism of the techniques 
involved. This is a tendency to be guarded 
against and one to which professional 
people and educators should give continued 
attention, particularly in the teaching of 
mathematics, science, and engineering. 
This paper, therefore, urges wider consider- 
ation of mathematical principles and meth- 
ods by scientists and engineers, whether 
or not they are primarily interested in 
mathematics, and gives examples with 
reference to one well-known mathematical 
concept. 2 

1. Introduction. — The concept of equiva- 
lence lies at the heart of nearly every trans- 
action, physical theory, and mathematical 
system. The literature abounds in such 
phrases as 

1 This paper was prepared from the manuscripts 
of two addresses by the writer: The concept (and 
misconcept) of equivalence, presented to the 
Philosophical Society of Washington on January 
18, 1947, and On the role of equivalence in pure and 
applied science and in practical and everyday life, 
presented to the Oberlin Mathematics Club, An- 
nual Banquet Meeting, on May 17, 1946. The 
writer wishes to express his appreciation of the 
many constructive suggestions and critical com- 
ments made during the preparation of the paper 
by his colleagues Dr. D. C. May, Mrs. A. B. 
McCaleb Nazary, and Miss A. Madsen, of the 
Bureau of Ordnance. The opinions expressed 
herein are those of the author and not necessarily 
those of the Navy Department. Received June 20, 

2 See also Burington, Richard S., New 
frontiers, Science 101: 313-320. Mar. 30, 1945. 


A is equivalent to B, 

which, unless properly denned, are often 
meaningless or misleading (e.g., "an elec- 
tric organ is equivalent to a pipe organ"; 
"one 1937 dollar is equivalent to 60 cents"). 

When S is a physical system, the ele- 
ments A, B, . . . of the system are identi- 
fied with specific physical objects, quan- 
tities, . . . , and the types of equivalence 
relations used must be carefully defined. 
Each such definition constitutes a separa- 
tion of A,B, . . . and the associated physical 
picture into classes. Such separation may be 
of considerable physical significance, each 
type of equivalence often having associated 
with it an extensive physical theory. Thus, 
in electrical theory are defined: "equiva- 
lent ra-pole networks," involving the "con- 
gruence" of certain matrices characterizing 
the physical system; "symmetric compo- 
nents," involving the "similarity" of certain 
matrices; etc. 

In the study of physical systems, perfect, 
or approximate isomorphic (abstractly 
identical) systems play a fundamental 
role. Equivalent physical (or mathematical) 
models are integral parts of the methods 
used (e.g., as in hydrodynamic similitude 
theories, . . . ). 

Considerable progress has been made by 
mathematicians in extending the theory of 
equivalence relations. To what extent 
these abstractions will be of value in the 
applied fields remains to be seen. This 
much is clear. A consciousness of the con- 
cept of equivalence is of real value. It 
helps to clarify problems. It provides a 
means of attack. The involved details of 
the modern abstract theories of equivalence 


will be in themselves of little value in prac- 
tical problems unless a penetrating inquiry 
into the economic, physical, or other perti- 
nent background of the problem at hand is 

2. The phrase (P) must be defined. — The 


A is equivalent to B, 

to have a well-defined meaning, requires 
appropriate definitions of the entities A and 
B, as well as a precise statement as to the 
meaning to be attached to the phrase "is 
equivalent to" as used in relation to A and 
B. Thus, the use of the phrase (P) should 
be accompanied by adequate and proper 
definitions. For example, in the phrase 
"1937 dollar is equivalent to 60 cents" 
it is not known what a "1937 dollar" is, 
what "60 cents" is, or what "is equivalent 
to" means. One would be at a loss to ex- 
plain what the phrase means. Likewise, 
the phrase "an electric organ is equivalent to 
a pipe organ" is questionable since there is 
nothing in the statement indicating in 
what sense the two organs are considered 
to be equivalent. Does the phrase mean 
that the tw r o organs can generate the same 
level of noise as measured (in some manner 
or other) in decibels in a given auditorium; 
does it mean that each of the two instru- 
ments can be used to play church music ; or 
does it mean that their tonal potentialities 
are comparable to each other, and so on? 
The list need not be extended further, for 
the reader can no doubt contribute hun- 
dreds of similar examples and queries. 

3. Importance of such statements as (P). — 
The importance of such statements as (P) 
becomes evident when one realizes that 
the notion of equivalence lies at the very 
basis of nearly every transaction and of 
nearly every mathematical system or form- 
ulation of a physical theory. For example, 
in order that the statement "3 feet = l 
yard" be understood, the concept of equiv- 
alent lengths must be thoroughly under- 
stood, and the fact that there exists stand- 
ards of length, such as 1 foot and 1 yard, 
must be appreciated. A proper understand- 
ing of such statements as "2X3 = 6" re- 
quires an accurate theory containing ade- 

quate definitions. Surely the operation of 
multiplying 2 by 3 is not "the same" as 6. 
If not "the same," then what is meant by 
the statement that 2X3 = 6? A complete 
answer to such questions requires a rather 
extensive treatment of the foundation of 
the theory of numbers. This will not be 
attempted here. 

4. Historical remarks on the definitions of 
equality. — As man first attempted to dis- 
cover the reasons for things, he began using 
the concepts of equivalence and identity. 
The philosophical considerations of these 
concepts have been debated at length. 
This phase of the subject is too vast to 
enter into here. 

Leibnitz, in Opera philosophica (Erd- 
mann) gave the definition: "Two things are 
called equal if, in every expression, one 
may be replaced by the other." This defini- 
tion is open to much criticism. 

Whitehead and Russell, in their Principia 
mathematica (1910), set forth a number of 
sets of postulates which an identity or an 
equality should satisfy. These postulates 
were developed over a period of time and 
were not original with Whitehead and Rus- 
sell. A certain set of these postulates is, es- 
sentially, as follows: 

I. Given any two elements A and B, either 
A = B or A 9^B. (Determinative property.) 
II. A = A. (Reflexivity property.) 

III. If A =B, then B=A. (Symmetrical prop- 

IV. If A =B and B = C, then A=C. (Transitiv- 
ity property.) 

The critical student might well object to 
the use of these postulates as the basis for 
definitions of identity or equality. However, 
they are a great improvement over earlier 

As MacDuffee 3 has pointed out, the 
teleological concept of equality has been 
favored by many mathematicians — until 
recently by most of them. (Teleology : the 
philosophical study- of evidence of a co- 
ordinated creative design in nature.) 

It should be noted that equality is some- 
times taken as one of the basic principles 
of logic and not subject to further definition 
or analysis. 

3 MacDuffee, C. C, Different kinds of equality, 
The Mathematics Teacher, Jan. 1936: 10-13. 

Jan. 15, 1948 

burington: the concept of equivalence 

The newer approach to the subject is to 
define a new type of equality for the ele- 
ments of some system. When this is done, 
it must be shown that the equality relation 
as defined actually has the required pro- 
perties I, . . . , IV. This removes the con- 
cept completely from philosophical study. 

5. Abstract definition of the phrase (P). — 
The abstract formulation of the notion of 
equivalence has been carefully studied by 
mathematicians. A brief outline of one 
such formulation seems appropriate for the 
purpose of this discussion, the formulation 
given being one commonly used in defining 
equivalence in algebraic systems, and in 
many formulations of physical theories. 

One abstract formulation of this concept 
of equivalence may be embodied in the fol- 
lowing four postulates and associated de- 

Let A, B, C, ... be elements in the fixed 
system S being considered (such as abstract 
entities, quantities, . . . ), and suppose a 
possible relationship that may occur be- 
tween these elements A, B, C, ... is de- 
fined in some manner or other. Suppose that 
when A and B are so related, the relation- 
ship is indicated by the symbol E, and 
the expression A E B is written; and that 
when A and B are not so related, the fact 
is indicated by the symbol A E B. Then 
the relationship expressed by the symbol 
A E B is known as the equivalence relation 
provided that it is denned to satisfy the 
following four properties : 

I. Determination. For any pair of elements A 
and B of S the relation A g B either holds 
or does not hold. 
II. Reflexivity. For any A one has A|i. 

III. Symmetry. When 4p, then B|i. 

IV. Transitivity. When AeB and BeC, then 
A EC. 

A is said to be equivalent to B if A E B is 
an equivalence relation. 

[The symbol E is to be read "is equiva- 
lent to"; the symbol Jp is to be read "is not 
equivalent to."] 

Every such definition of equivalence, 
which is ordinarily not Unique for the given 
system, constitutes a division of the ele- 
ments A, B, C, . . . into classes. 

The determinative property (I) suggests 
the existence of at most two cases for the 

definition of equivalence used, and that the 
relation is determinative. 

The reflexive property (II) is an exten- 
sion of the earlier equivalence relation 
known as the identity. 

The symmetric property (III) insures 
that the relation of equivalence is sym- 

The transitive property (IV) is an ex- 
tension of the old concept "things equal to* 
the same thing are equal to each other." 

The special equivalence relation A V B, 
which is defined to hold for any pair of 
elements A and B is called the universal 

The equivalence relation A I B, which 
holds only when A and B are the same or 
"identical elements," is called the identity 
or unit relation. 

A serious study of systems which satisfy 
these properties would be a large under- 
taking. It would include study of systems 
that do not satisfy all these- properties. 
Such an undertaking would involve a great 
many fields of mathematics. This will not 
be attempted in the present paper. 

6. Equality and equivalence. — It should 
be remarked that the formulation given in 
paragraph 5 has been used to define the 


A is equal to B. 

For the purposes of this paper the fine 
points of reasoning which have led some 
scholars to use the above postulates to 
define "is equal to" and others to use them 
to define "is equivalent to," and to dis- 
tinguish between "is equal to" and "is 
equivalent to," need not be discussed here. 

7. Ordinary plane (Euclidean) geometry. 
— Ordinary plane geometry when viewed 
from the newer point of view is crudely as 
follows : 

(1) The elements of this geometry are points 
and lines forming geometric figures. 

(2) The concept of superposition is taken for 
granted in such an intuitional geometry. 

(3) Two geometric figures A and B are called 
congruent (equal) if one figure may be ro- 
tated and translated until it is brought into 
coincidence with the other figure. 

(4) Given any two geometric figures A and B : 
(a) Either these figures can be made to co- 
incide (A E B), or they can not be made 


VOL. 38, NO. 1 

to coincide (A E B). Thus the property 
of determination holds for the geome- 
try under consideration. 

(b) Every figure can be superimposed upon 
itself. (4 E A). Hence the property of 
rcflexivity holds. 

(c) If one figure can be made to coincide 
with a second figure, (A E B), then the 
second can be made to coincide with 
the first, (B E A). Thus the property of 
symmetry holds. 

(d) If one figure can be made to coincide 
with a second figure, (A E B), and the 
second figure can be made to coincide 
with a third figure, (B E C), then the 
first can be made to coincide with the 
third, (i E C). Hence, the property of 
transitivity holds. 

Therefore, ordinary plane (Euclidean) 
geometry satisfies the four required pro- 
perties of an equivalence relation as de- 
fined in paragraph 5. 

(5) In this type of geometry, the geometry of 
congruent figures is the study of those prop- 
erties relating to the figures which remain 
invariant under the equivalence relation of 
congruence. (Thus, if two triangles are con- 
gruent, their areas are equal.) 

8. Other examples from elementary ge- 
ometry. — In addition to the example given 
above there are many other simple exam- 
ples of equivalence to be found in elemen- 
tary geometry. Thus, in Euclidean geo- 
metry, two polygons A and B may be 
equivalent in the sense that they have 
equal areas, but they may or may not be 
equivalent in the sense that they have the 
same number of sides or angles, and so on. 

Again, any two proper conies, A and B y 
are equivalent (that is A E B) in the sense 
that A may be transformed into B by 
means of a projective transformation, 
while on the other hand A may or may not 
be equivalent to B in the sense that A may 
be transformed into B by means of a trans- 
lation and rotation. 

In general, in geometry, whether two 
geometric quantities are equivalent can be 
determined by calculating certain numbers 
known as invariants, which are associated 
with the objects under study. In such cases, 
if certain relative invariants for an object 
A are identical, respectively, to the cor- 
responding invariants for the object B, 
the object A is equivalent to the object 
B. If these two sets of invariants are not 

identical, respectively, A is not equivalent 

For example, in ordinary translational 
geometry, two lines 


ax-\-by-\-c = 0, 



dx+ey+f = 0, 

are equivalent in the sense of being parallel 
if, and only if, the ratio a/b is equal to the 
ratio d/e. These ratios are invariants of the 
lines (1) and (2), respectively, for a trans- 
lational geometry. 

Ordinary plane similarity geometry. The 
geometry of similar figures is a geometry 
quite distinct from ordinary Euclidean 
geometry. In this geometry the relation of 
similarity is easily shown to satisfy all 
four properties of an equivalence relation. 

Other geometries. There are many other 
geometries, and the relationships of con- 
gruence used in them are examples of 
equivalence relations. Familiar examples of 
these types of geometries are projective, 
non-Euclidean, affine, and the like. 

9. Example from transportation. — In 
order to show how these equivalence prob- 
lems come up in practical fields outside of 
mathematical subjects, an example may be 
noted in the field of railroading. In the 
transportation literature many types of 
equivalence are commonly used. An ex- 
ample of current discussion in the trans- 
portation literature 4 is that of locomotive 
ratings, "equivalent locomotives." A num- 
ber of methods of rating locomotives are 
commonly used, and much general con- 
fusion exists as the result of careless use of 
various types of equivalence relations. 

Under present conventional systems of 
rating, a steam, a Diesel, and an electric 
locomotive, each locomotive rated as hav- 
ing "a 6,000 h.p. output," are actually only 
equivalent in the sense that the number 
6,000 used happens to be the same, since 
the method of calculating the horsepower 
number is totally different in each case 
(one is cylinder output; the second is 
Diesel-engine output; the third is continu- 

4 Wynne, F. E., Comparable locomotive ratings, 
Railway Age 120 (6): 316-318. Feb. 9, 1946. 

Jan. 15, 1948 

burington: the concept of equivalence 

ous output at rails; and each of them re- 
quires careful definitions). 

If three locomotives are equivalent in the 
sense that they can each deliver the same 
useful output to the rail, they are not neces- 
sarily equivalent in the sense of their ability 
to handle the same weight of cars at 100 
m.p.h. on level tangent track. And, if 
three locomotives are equivalent in this 
latter sense, they are not necessarily equiv- 
alent in the sense that they can deliver 
the same useful horsepower continuously at 
the rails; nor are they necessarily equivalent 
in the sense of their earning power, avail- 
ability, reliability, etc. 

Actually under present conventional 
systems of ratings, three locomotives hav- 
ing an advertised 6,000 h.p. output, one 
steam, one Diesel, one electric, while 
equivalent in the sense that the number of 
6,000 is the same for all, are not equivalent 
in the sense of the weight of cars that they 
can handle at 100 m.p.h. on level tangent 
track continuously, the values actually be- 
ing about 1,000 tons, 900 tons, 1,300 tons, 
respectively. Nor are they necessarily 
equivalent in their cost per ton mile hauled, 
availability ratio, etc. Furthermore, there 
are many other equivalence relations used 
in rating locomotives, all different, and 
many times badly misunderstood because 
of the lack, or omission, of good defini- 
tions. A general practical definition of 
"equivalent locomotives" has never been 
adequately given; and any such definition 
would doubtlessly involve the listing of 
many categories of equivalence and a 
scheme, or a set of schemes, for weighting 
these equivalences. A similar situation 
exists in many other fields. 

10. Equivalence in the elementary theory 
of sets. — In the elementary theory of sets, 
one of the basic notions is that of equiva- 
lence. If the elements in two sets, A and B, 
can be paired with each other in such a 
manner that to each element of A there 
corresponds one and only one element of 
B, and to each element of B there corre- 
sponds one and only one element of A, then 
the correspondence is said to be bi-unique, 
and A and B are said to be equivalent. 

Two finite sets have the same number 
of elements if and only if the elements of the 

two sets can be put into bi-unique corre- 
spondence. This is the idea of counting, for 
when one counts a finite set of objects one 
simply establishes a bi-unique correspond- 
ence between these objects and a set of 
number symbols 1, 2, 3, . . . , n. Thus, the 
notion of equivalence for finite sets corre- 
sponds to the ordinary notion of equality of 

The concept of equivalence has been 
extended to infinite sets. This was done to 
construct an arithmetic of infinities. In this 
sort of theory there are just as many points 
on a straight line as there are real num- 
bers. This means that the set of all real 
numbers and the set of all the points on a 
straight line are equivalent in the sense 
that, once an origin and a unit are chosen, 
a bi-unique correspondence between the 
real numbers and the points on the line 
can be made. With this understanding of 
equivalence, a finite set cannot be equiva- 
lent to any one of its proper subsets, for 
if the finite set contains n elements and 
no more, any one of its proper subsets can 
contain at most n — 1 elements. 

If a set contains infinitely many objects, 
it may be equivalent to a proper subset of 
itself. For example, there are just as many 
positive integers as there are positive even 
integers. This is easy to see from the bi- 
unique correspondence shown below. 

12 3 4 5- 

• • n • 

1 I J t t 


2 4 6 8 10 • 

• • In 

In fact, there are just as many rational 
fractions as there are integers. However, 
the set of all real numbers is not equivalent 
to the set of integers. 

For those who wish to pursue this sort 
of equivalence theory, much can be found 
concerning the subject in the theory of 
sets (begun by George Cantor at the end 
of the nineteenth century) . 

11. Algebra. — Algebra furnishes many 
illustrations of the appropriateness of the 
concept of equivalence. Thus, if real posi- 
tive numbers are assumed to be properly 
defined, then it is possible to introduce 
negative numbers in quite a logical manner 
by means of a certain definition for the 
equivalence of pairs of real positive num- 



bers. Such a treatment may be found in the 
literature 5 and gives quite satisfactory an- 
swers to such questions as: Why does 
( — !)(— 1) = 1? A similar treatment based 
on the equality of pairs of real numbers can 
be formulated to give a logical introduction 
to complex numbers. 

12. Klein 1 s definition of a geometry. 
Groups. — In order to illustrate the connec- 
tion between the theory of groups and the 
theory of equivalence further examples 
from geometry may be considered. Felix 
Klein (1871) defined geometry as the study 
of the invariants of a group of transforma- 
tions. A group of transformations is a set 
such that: 

1. The resultant transformation of two trans- 
formations is a transformation of the set. 

2. The associative law holds. 

3. There exists an identity transformation. 

4. Every transformation has an inverse. 

In ordinary Euclidean geometry, the 
transformations (of rotations and transla- 
tions) form a group. Likewise in similarity 
geometry, the similarity transformations 
form a group. Thus, the geometry of con- 
gruent figures and the geometry of similar 
figures are examples of geometries as de- 
fined above. 

13. The resemblance between the definition 
of group and the definition of equivalence. — 
This resemblance is a fundamental one. 
In associating an equivalence relation with 
a set of transformations, one associates 
with a set A of elements the set T of trans- 
formations which operates on A transform- 
ing it into some set B of elements. One 
might make a definition of equivalence by 
saying that B is equivalent to A if there 
exists a tranformation in the set T which 
transforms A into B, provided of course 
that the four properties of an equivalence 
relation are met. That this can be done, if 
and only if the transformations T form a 
group, can be shown readily. 

It can be shown quite easily that: 

Theorem. Equivalence relative to a group of 
transformations is an equivalence relation 
(satisfying properties I, II, III, IV). 

Something of what this theorem means 
may be illustrated by the following example : 

5 See C. C. MacDuffee, loc. cit. 

Example. Consider the set of all real rotations 
T of the points in a plane about a fixed point 0. 
This group may be represented by the equation 

w=ze i6 . {%*■ 



By this formula any point z in the plane having 
polar coordinates p and a is rotated about the 
point into a point w in the plane having polar 
coordinates p and {a +0). 

That the set (7') of rotations form a group can 
be seen readily since: 

1. The resultant of two rotations 0i and 02 is a 
rotation 0=0!+0 2 of the set (T). 

2. The associative relation holds since for suc- 
cessive rotations, 6 h 2 , 03, 0i + (02+03) 
= (0i +0 2 ) +03- 

3. The rotation =0 is the identity transforma- 

4. Every rotation has the inverse rotation 

All points P in the plane at a fixed distance p 
from the point could then be defined as equiv- 
alent with respect to the group of rotations (T) in 
the sense that any one point A of the points P can 
be transformed into any other point B of the set 
P by an appropriate rotation 0. 

Thus, in this sense, all points on a circle with 
radius p and center are equivalent to each 
other. But points on this circle are not equivalent 
to points on a circle with center at and radius r 
where r i£p. However, all points on the second 
circle are equivalent to each other in the sense 

That this definition satisfies the four properties 
for an equivalence relation follows since: 

(A) Either two points in the plane fall on the 
same circle with center at 0, or they do 
not. Thus property (I) is satisfied. 

(B) Two coincident points lie on the same circle 
with center at 0; hence property (II) is 

(C) When one point is equivalent to a second 
point they fall on the same circle with 
center at 0; hence the second point is 
equivalent to the first point. Thus prop- 
erty (III) holds. 

(D) When a point A is equivalent to a point B, 
they lie on a circle with center at 0, and if 
B is equivalent to C they lie on a circle 
with center at 0; hence A and C lie on the 
same circle since B lies on both circles with 
center at 0. Thus property (IV) holds. 

This example serves to illustrate the 
resemblance between the definitions of 
group and equivalence. 

14. Isomorphic systems. — Another math- 
ematical concept of considerable impor- 
tance in its own right as well as in its appli- 
cations is that known as isomorphic sys- 

Consider two systems A and B each 
consisting of a set of elements and a set of 

Jan. 15, 1948 

burington: the concept of equivalence 

operations on these elements. Suppose that 
each system is closed with respect to a given 
system of operations in the system. The 
two S3^stems A and B are said to be iso- 
morphic or abstractly identical with respect 
to these operations if there exists a bi- 
unique one-to-one correspondence between 
the elements of A and B, such that any 
formal combination of, or operation on, 
the elements in A corresponds to the analo- 
gous construction with the corresponding 
elements in B. 

In mathematics (such as in algebra) 
two isomorphic systems are commonly 
considered as equivalent; and the subject 
matter of mathematics, in this sense, may 
be considered as dealing with those pro- 
perties of systems which are invariant 
(remain unchanged) for isomorphic sys- 

15. Applications to physical phenomena. 
— In any specific situation the elements 
a, b, c ... of a system S are identified with 
specific physical objects, numbers, quanti- 
ties, entities, or the like, and the definitions 
of equivalence used must be carefully given 
in terms of these elements and the physical 
systems to which they belong. Each defini- 
tion of equivalence used constitutes a 
separation of the set of elements a,b, c, . . . 
and associated physical phenomena or 
systems into classes. This separation into 
classes is often of considerable physical 
significance, each type of equivalence often 
having associated with it an extensive 
physical theory. 

16. Examples from circuit theory. — As in 
mathematical theories, so in physical 
theories, many non-isomorphic types of 
equality have been and can be defined, 
each type of equality often having associ- 
ated with it an extensive theory. Thus, in 
the theory of equivalent linear electrical 
circuits, two 2-pole networks may be equiv- 
alent 6 in the sense that for all frequencies 
they have identical driving-point admit- 
tances (or more generally, for 2N-pole net- 

6 Buhington, Richard S., Matrices in electric 
circuit theory, Journ. Math, and Phys. 14 (4): 
325-349. Dec. 1935; A matric theory development 
of the Theory of Symmetric Components, Philos. 
Mag. (ser. 7) 27: 605. May 1939; On circavariant 
matrices and circa-equivalent networks, Trans. 
Amer. Math. Soc. 48 (3): 377-390. Nov. 1940. 

works, that they have identical charac- 
teristic coefficient admittance matrices), 
yet they may not be equivalent in the sense 
that they have the same number of inde- 
pendent mesh circuits; or, are structurally 
the same; or, are equally economical to 
operate; or, are both readily physically 
realizable ; and that if A and B respectively, 
are their network matrices, A and B may 
or may not be equivalent in the sense of 
matric congruence, . . . 

Again in the theory of symmetric com- 
ponents as used in electrical engineering, 
the equivalence relationship used may 
often involve that of matric similarity be- 
tween the matrices used to represent certain 
characteristics of the network in the various 
reference systems used in the theory. In 
this sort of equivalence the actual values 
of the voltages, currents, and impedances 
are left undisturbed, though the values of 
their representations in the various refer- 
ence systems may be greatly different. Here 
the equivalence relation known as matric 
similarity is distinctly different from the 
equivalence relation known as matric con- 
gruence. Yet, both of these types of equi- 
valence happen to be examples of another 
type of equivalence known as ordinary 
matric equivalence. 

17. Isomorphism as used in model stud- 
ies. Principles of similitude. — In the design 
of structures, bridges, ships, dams, flood 
control projects, and the like, the engineer, 
naval architect, and others responsible for 
the design must make accurate predictions 
as to the characteristics, cost, and perform- 
ance of the various proposed designs. In 
such work the designers and planners can 
ill afford to make errors. Such projects are 
too expensive. Perhaps only one can ever 
be constructed. The final product must be 
right. It must do what it is designed to do, 
reliably, safely, and economically. 

Engineers and scientists, in such instan- 
ces, frequently take recourse to the con- 
struction and testing of models of the pro- 
posed structure. The results of the tests of 
the models are then used to predict the per- 
formance and characteristics of the pro- 
posed prototype. In order that a model test 
be of real use in predicting the properties of 
the prototype, great care must be exercised 



in its design. Furthermore, considerable 
care is necessary in interpreting the meas- 
ured and computed properties of the model 
test in terms of the prototype. 

Theoretically, in designing a model test 
and in constructing the models for the 
test, every physical parameter of impor- 
tance in the prototype must be considered 
and taken into account. It is not sufficient 
merely to make the dimensions of the model 
and prototype proportional. If a theory of 
similitude is not available, then one must 
be developed. Such a theory must serve as 
the basis for designing and constructing the 
model from its prototype dimensions. It 
must serve also as the guiding principle in 
interpreting the measurements made on the 
model in terms of the corresponding meas- 
urements as predicted for the prototype. If 
this theory of similitude is a perfect one, 
and the application of this theory is also 
perfect, then there is a one-to-one bi-unique 
correspondence between the various physi- 
cal parameters of the prototype and the 
corresponding ones of the model ; and every 
operation on or with the prototype has a 
corresponding operation on the model. In 
other words, any characteristic of the model 
as predicted through the isomorphism thus 
stated must be a true characteristic of the 
prototype, and vice versa. (This is an ex- 
ample of the equivalence property, if A E 
B, then BE A.) 

In practice it is impossible to apply per- 
fectly such a theory and expect the results 
predicted through the model test to be 
completely true for the prototype. A meas- 
ure of this perfection lies in a comparison of 
the actual characteristics of the prototype 
with those predicted by the model studies. 
Consequently the aim in all such work is 
the construction of as near perfect an equiv- 
alent system of models as is humanly pos- 
sible. In other words, these models must be 
as nearly isomorphic with their correspond- 
ing prototypes as possible. 

An example where consideration is given 
to many physical parameters of importance 
may be found in hydrodynamics. In hydro- 
dynamic studies, lengths L, a, b, c, . . . , 
time T, velocity V, mass M, force F, 
pressure increment p, mass density p, 
specific weight 7, viscosity n, surface 

tension c, and clastic modulus e for the 
object and fluid must be considered. (For 
convenience, L, M, F may be taken as the 
three fundamental dimensions.) 

A number of theories of similitude have 
been developed for use in aero- and hydro- 
dynamics. In one such theory, widely used, 
there is defined a certain set of dimension- 
less numbers 


7T 6 = 


7T 4 

7T 7 

P 7 2 

V 2 a 


7T 5 

V 2 /a 


7T8 = 


which must be kept unchanged if true 
similarity (i.e., isomorphism) is to exist 
between the flow about the prototype and 
the flow about the model. In other words, 
if true similarity is to exist, every dimen- 
sionless parameter x n , referring to condi- 
tions in the model, must have the same 
numerical value as the corresponding 
parameter for the prototype. This means, 
for one thing, that the model and prototype 
must be completely similar geometrically. 
To put the problem in another way, the set 
of dimensionless numbers ari, . . . for the 
prototype must be identical with the cor- 
responding set xi, . . . for the model — if 
true similarity between prototype and 
model exists, and if predictions made from 
model studies are to be valid for the actual 
prototype. A mathematician would say that 
the parameters in . . . , must be absolute 
invariants for the prototype and model 
systems. Because the quantities xi, . . . , 
ir n , . • . are invariant for these isomorphic 
systems, the model and prototype systems 
are said to be equivalent. 

Thus, if wi = a/b is to be invariant, where 
a and b are any two linear dimensions of the 
prototype, the corresponding linear di- 
mensions a! and b' of the model must be so 
related that iri = a'/b'. The reader can, for 
himself, discover other requirements on the 
model by merely holding each of the other 
parameters 7r 2 , . . . fixed and interpreting 
the quantities in these parameters first, in 
terms of the prototype, and secondly, in 
terms of the model. 

From a practical standpoint it is usually 

Jan. 15, 1948 

burington: the concept of equivalence 


impossible to realize fluids and values of the 
physical parameters of these fluids to 
satisfy all the requirements implied when 
the absolute in variance of the set in, . . . is 
demanded. This means that it is impossible 
to obtain a true model on any but the same 
scale using any but the same fluid as proto- 
type. Of course, in this case, the model and 
the prototype would be equivalent. (This is 
an example of the equivalence property 
A El) In spite of the difficulty and the 
knowledge that any practical model system 
cannot be made perfectly isomorphic with 
its prototype, much can be learned and 
reasonably reliable predictions can be ob- 
tained from such model studies. 

In the testing of ship models and partly 
submerged objects, such as buoys and sea- 
planes, it has been found possible to make 
rather good predictions as to the behavior 
of the prototype by designing the model 
studies so as to keep the Froude number ts 
invariant. This compromise places the 
emphasis on the dominant physical para- 
meters involved, namely, inertia and gra- 
vity forces, since 7r 6 is their ratio. In such 
work, 7T4 known as Newton's number, can 
also be kept invariant. As a rule it is not 
possible to keep the remaining functions 
7T2, T3, 7T 6 , . . . invariant. This compromise 
means that the model system will be equiva- 
lent to the prototype system in the sense 
that the Froude and Newton's numbers are 
identical, respectively, for both the proto- 
type and the model; but it does not mean 
that the prototype system and the model 
system are completely isomorphic other- 

When other physical parameters different 
from mass and inertia are considered to be 
of greater importance, then some other 
dimensionless numbers become of prime 
interest, and the functions in, ttj, correspond- 
ing to the two most important physical 
parameters are made absolutely invariant. 
Thus, in aerodynamics, emphasis is often 
placed on Reynold's number, ir%, rather 
than the Froude number x 5 , since skin fric- 
tion (viscosity) is then of prime importance. 
T6 is the ratio of inertia and viscous forces. 

Where inertia force and compressibility 
predominate, emphasis is placed on the 
Mach number, 7r 8 , which is the ratio of the 

velocity of flow to the velocity of sound in 
the fluid at the given temperature. 

18. The use of equivalence principles in 
the design and testing of equipment. — Air- 
craft, railway cars, and the like must be able 
to withstand great stresses and strains and 
much rough handling and must not be too 
vulnerable to damage from a great variety 
of causes. In designing such equipment a 
great deal of attention must be given to 
strength, safety, reliability, and costs, while 
recognizing economical and utilitarian 
values. Such considerations commonly in- 
volve various types of testing procedures 
designed to indicate the strength, reliabili- 
ty, etc, of vital portions of the equipment, 
as well as of the structure as a whole. Thus, 
the wings of an airplane or the axles of a 
truck can be tested in various ways for 
strength, ability to withstand shock, etc; 
the vulnerability of the fuselage or body to 
fire can be studied; and so on. But the re- 
sults of such studies are only a partial indi- 
cation of the strength and vulnerability of 
the airplane or car as a whole. Conse- 
quently, when possible, tests of the equip- 
ment as a whole are sometimes set up and the 
results obtained compared with the results 
of the tests of specific parts' of the equip- 
ment. Of course, the ultimate test lies in the 
experience gained with the equipment un- 
der actual service conditions for a long pe- 
riod of time under a great variety of cir- 

The quantities involved in criteria de- 
veloped for use in testing may be quite 
different from the quantities available from 
over-all testing or from operational experi- 
ence. An adequate theory for correlating 
(a theory of equivalence) these measures 
obtained in tests of specific parts and in 
over-all testing and experience must be 
formulated. Thus in attempting to measure, 
say, the ability of a car to withstand colli- 
sion, controlled tests might be made in the 
laboratory, in which such parameters as 
energy, velocity, deformation, stress distri- 
bution, bending moments, momentum, 
pressure, and the like are used. Yet such 
parameters as these may not be available 
in examining the damage to such a car in 
collision ; the only real information available 
being that which can be observed and de- 



VOL. 38, NO. 1 

duced from the wreckage. The need for a 
theory for correlation (equivalence) is evi- 
dent in such a case. 

To summarize in connection with such 
test results there appears: 

(1) The problem of the formulation of criteria 
and methods for describing, defining, and 
measuring such things as strength, safety, 
vulnerability to damage, ... of the com- 
ponent parts of the equipment, as well as of 
the equipment as a whole. 

(2) The problem of formulating principles for 
correlating these measures of strength, 
safety, damage, .... These measures may 
be observed in experiments — 

(a) with specific parts of the equipment, 
(6) with specific assemblies of the equip- 

(c) with the equipment as a whole under 
test conditions, 

(d) with the equipment as used in actual 

In order that an adequate basis for 
reliably predicting the worth and safety of 
a design be realized, each of these major 
problems must be faced and an adequate 
solution obtained when at all possible. The 
construction of such a theory and mode of 
prediction, if it is to be a good one, must 
involve a careful use of the principles of 
equivalence. The importance of this pro- 
cedure is all the more important since ac- 
tual operational experience with a new de- 
sign of certain types of equipment may be, 
costly (or even dangerous to human life) 
and such experience cannot always be made 
available before production. 

Thus, in the field of design, testing, etc., 
there is a continuing need for the wise use 
of the principles of equivalence. 

19. Recent mathematical developments 7 . — 
In recent years much progress has been 
made by such mathematicians as Ore, 
MacDuffee, Garrett Birkhoff, and others in 
extending the analysis of the theory of 
equivalence relations. These extensions 
have served to connect rather diverse math- 
ematical fields and have gone deeply into 

7 Whitman, P. M., Lattices, equivalence rela- 
tions and subgroups, Bull. Amer. Math. Soc. 
52 (6): 507-522. June 1946; Ore, Oystein, 
Theory of equivalence relations, Duke Math. 
Journ. 9 (3). Sept. 1942; Birkhoff, Garrett, 
On the structure of abstract algebras, Proc. Cam- 
bridge Philos. Soc. 31: 433-454. 1935; Mac- 
Duffee, C. C, loc. cit. 

various fields of abstract algebra, topology, 
and related fields and apparently are lead- 
ing to still a more general and abstract 
theory of mathematical relations. These 
investigations are of interest chiefly to 
workers in certain branches of pure mathe- 
matics. To what extent some of these ab- 
stractions will be of immediate value to 
workers in the applied fields remains to be 
seen. But this much seems clear, that: 
Consciousness of the concept of equivalence 
as outlined herein is of real value in many 
physical fields. It helps to clarify many 
problems. It provides a means of attack. It 
should be remarked, however, that the 
involved details in the development of the 
modern theory of equivalence relations 
will in themselves be of little value in prac- 
tical and scientific problems unless a pene- 
trating inquiry into the economic, physical, 
or other pertinent background of the prob- 
lem at hand is made. This last point is of 
great importance. 

To illustrate what is meant, consider the 
following well known mathematical results: 

A partition P of the set S is a decomposition 
of S into subsets Ci, • • • , C„, • • • such that every 
element in S belongs to one and only one set C n . 
The sets C n are called blocks of the partition P, 
and P = P(C n ) is written to indicate this. 

Theorem. Any partition P(C n ) defines an 
equivalence relation E in the set S when one 
puts AeB whenever A and B belong to the 
same block C n . Conversely, any equivalence 
relation E defines a partition P(C n ) where 
the block C n consists of all elements equiva- 
lent to any given element A . 

This theorem tells us that there are many 
possible equivalence relations definable for 
a set S. Which particular equivalence rela- 
tions are worth studying seriously will de- 
pend on the set itself and what general 
problems are under consideration. Thus, if 
S is the set of all locomotives, many types 
of equivalance relations can be defined, such 
as equivalence in the sense of the same 
cylinder output ; or continuous output at the 
rails; or equivalence in the sense that thej^ 
have, or do not have, brass trimming around 
the edge of the headlight; or equivalence in 
the sense that they can reach a maximum 
speed of 100 miles per hour, or not; or 
equivalence in the sense that the locomo- 

Jan. 15, 1948 Hermann: miscellaneous middle American lonchocarpi 


tives are named after some one, or are not; 
etc. Now obviously, some of these types of 
equivalence have some reason for existence, 
while others are of course trivial or ridicu- 
lous. To repeat the point: Involved details 
in the theory of equivalence relations will 
in themselves be of little value in practical 
problems unless a penetrating inquiry into 
the economic, physical, or other pertinent 
background of the problem at hand is made. 
20. Summary. — The present paper dis- 
cusses some of the meanings that may be 
attached to the phrase "A is equivalent to 
B." A glimpse of the mathematical prop- 
erties that must be possessed by an 
equivalence relation has been shown. Iso- 
lated examples of equivalences in geom- 
etry, set theory, number theory, algebra, 
electrical networks, hydrodynamics, and 
engineering have been cited. Some atten- 
tion has been given to the theories of 
modeling and similitude, which are so im- 
portant in hydro- and aerodynamics, 
theories in which a form of equivalence 
known as isomorphism plays a leading role. 

The use of either perfect or approximate 
isomorphic systems (or more general equiv- 
alent systems) appears as a fundamental 
process in almost all studies of physical 
phenomena. The method consists broadly 

(1) The extraction from the physical phe- 
nomena S of a nearly isomorphic (equiv- 
alent) physical model P. 

(2) Reduction of the physical model P to 
an isomorphic (equivalent) mathematical 
model M amenable to treatment. 

(3) A solution of this mathematical system M. 

(4) The interpretation of the solution found in 

(3) in terms of the mathematical model M. 

(5) The interpretation of the solution found in 

(4) in the physical model P. 

(6) Finally, the interpretation of the result (5) 
in the original physical settings. 

Although at present it is not clear how 
much value the abstract extensions of the 
theory of equivalence relations will prove 
to be for use in the applied fields, it does 
appear that the theory will be beneficial in 
offering a background for the broad ap- 
proaches to practical problems. 

BOTANY.- — Studies in Lonchocarpus and related genera, II: Miscellaneous Middle 
American Lonchocarpi. 1 Frederick J. Hermann, U. S. Department of 

The most extensive and generally useful 
of the comparatively recent partial treat- 
ments of the genus Lonchocarpus is Henri 
Pittier's The Middle American species of 
Lonchocarpus (Contr. U. S. Nat. Herb. 20: 
37-93. 1917). This monographic account 
embraces the 40 species known from Mexico 
and Central America 30 years ago, to which 
are appended a list of six excluded or doubt- 
ful species and detailed descriptions of nine 
related South American and West Indian 
Lonchocarpi. One of the first tasks of a 
current review of the genus as a whole is, 
therefore, to attempt to allocate within the 
framework of the classification proposed by 
the author of that work the miscellaneous 
species subsequently described by various 
other authors from the same area. In some 
cases the systematic position of a recently 
proposed species has been correctly indi- 
cated by its author and characteristics dis- 

1 Received July 8, 1947. 

tinguishing it from its nearest allies may have 
been pointed out; in others lack of either 
flowering or fruiting material may have 
prevented this; in still others a misinterpre- 
tation, due either to faulty earlier descrip- 
tions or to the author's lack of comprehen- 
sive familiarity with the group, may vitiate 
the supposed relationship and hence sec- 
tional position or taxonomic status; and, 
finally, in some instances no attempt what- 
ever has been made to indicate the rela- 
tionship of the new species. 

No discussion seems to be required here 
of such of these species as have been satis- 
factorily disposed of by other authors, such 
as Lonchocarpus caribaeus Urban (referred 
to the synonymy of L. benthamianus Pit- 
tier by Harms in Fedde Rep. Spec. Nov. 
17: 323. 1924), L. capensis M. E. Jones 
(shown to be actually Tamarindus indica 
L. by Morton in Contr. U. S. Nat. Herb. 29 : 
103. 1945), L. modestus Standi. & Steyerm. 
(transferred to Lennea by its authors in 



Fieldiana, Botany, 24 (5): 275. 1946), and 
L. trifoliolatus Standi, (equated with L. 
phaseolifolius Benth. by Standley and 
Steyermark in Fieldiana, Botany, 24 (5): 
282. 1946). Six additional names {L. argy- 
rotrichus Harms, L. calderoni Standi., L. 
lindsayi Standi., L. obovatus Benth., L. 
salvinii Harms, and L. schiedeanus (Sch- 
lecht.) Harms) have been recently trans- 
ferred to Willardia by the writer (Journ. 
Washington Acad. Sci. 37: 427. 1947), 
to which genus Standley (Contr. U. S. 
Nat. Herb. 23: 483. 1922) had already 
referred L. eriophyllus Benth. 

It seems desirable to present the conclu- 
sions of the writer upon the following ad- 
ditional eight species, particularly since 
certain original misinterpretations are 
being perpetuated in recent important 
floristic works such as the Flora of Guate- 

Lonchocarpus amarus Standi., Carnegie 
Inst. Washington Publ. 461: 63. 1935 = Va- 
tairea lundellii (Standi.) Killip ex Record 
{Tipuana lundellii Standi. I.e. 65). 

Since the original material of Tipuana lun- 
dellii lacked flowers and that of Lonchocarpus 
amarus is without fruit, the failure to realize 
the identity of the two is readily understand- 
able. The alternate leaflets, diadelphous sta- 
mens, wings of the corolla free from the keel, 
and the strikingly funnelform calyx of Loncho- 
carpus amarus definitely exclude it from that 
genus. Flowering specimens of Vatairea lun- 
dellii (collected since the publication of that 
species), kindly lent to the writer for study by 
the Chicago Natural History Museum, are 
identical with the type of L. amarus. Vatairea 
lundellii appears to be most nearly related to 
the Amazonian V. fusca Ducke, from which it 
is distinguishable by the appressed rather 
than spreading pubescence of the calyx, pedi- 
cels, and peduncles, by its broader wing petals 
(4.5 mm), by having the stamens definitely 
shorter than the pistil, and by the tendency of 
the calyx to split in the late-bud stage, be- 
tween the vexillar teeth. 

Lonchocarpus dumetorum Brandegee, Univ. 
California Publ. Bot. 10: 181. 1922 = L. dari- 
ensis Pittier, Contr. U. S. Nat. Herb. 20: 69. 

The type specimen of L. dumetorum (Purpus 

8591 ; this and other Brandegee types reviewed 
through the courtesy of the University of 
California Herbarium) is a good match with 
that of L. dariensis {Pittier 5515, U. S. National 
Herbarium). It has, in addition to flowers and 
leaves, nearly mature legumes, which are very 
closely similar to those of L. megalanthus 
Pittier, this similarity bearing out Dr. Pittier's 
tentative alignment of it, in the absence of 
fruit, with the latter species. Both L. dariensis 
and L. megalanthus, however, as well as the 
closely related L. mexicanus Pittier, have leaflets 
conspicuously punctate, and so they could not 
be keyed out to his section Epunctati wherein 
he placed them. It seems probable that this is 
what led Brandegee to believe that he had an 
undescribed species in his L. dumetorum, and 
the same discrepancy may have been partly 
responsible for his proposal of L. purpusii. 

Lonchocarpus izabalanus Blake, Contr. U. 
S. Nat. Herb. 24: 7. 1922 = L. luteomacula- 
ttjs Pittier, Contr. U. S. Nat. Herb. 20: 64. 

Lonchocarpus luteomaculatus is a highly vari- 
able species, almost as polymorphic in fact as 
its near ally L. latifolius (Willd.) HBK. The 
type of L. izabalanus is very similar to many 
recent collections of L. luteomaculatus. It was 
differentiated, in the original description, prin- 
cipally bj' the possession of a maroon banner 
with a green spot at the base instead of a 
purple banner with a basal yellow spot, but it 
seems likely that the flower color ascribed to 
L. luteomaculatus by its author was that of the 
dried plant and that in the fresh state it may 
be actually closer to the maroon-green pattern. 
At any rate, there now seems to be no dis- 
tinguishable difference between the corolla 
colors of the two t} r pe specimens {Blake 784-1 
and Pittier 4170, U. S. Nat. Herb.). 

Lonchocarpus kerberi Harms, Fedde Rep. 
Spec. Nov. 17: 322. 1921 =L. peninsularis 
(Donn. Smith) Pittier, Contr. U. S. Nat. 
Herb. 20: 56. 1917. 

A fragment of the type of L. kerberi {Kerber 
35) is fortunately preserved in the herbarium 
of the Chicago Natural History Museum. This 
is sufficiently ample to show that it is not at all 
related to L. benthamianus Pittier, L. proteran- 
thus Pittier, and L. punctatus HBK., as sup- 
posed by Harms, and that it differs from the 
type of L. peninsularis {Tonduz s.n. {Inst. 

Jan. 15, 1948 Hermann: miscellaneous middle American lonchocarpi 


Fis. Geogr. Costa Rica 13961), U. S. Nat. Herb.) 
only in the corolla being slightly less pubescent. 
It is not surprising that L. kerberi was regarded 
as a new species by its author and that its 
affinities were misinterpreted. Since it has 
leaflets that are not impressed-nerved and 
that are unmistakably punctate, its relation- 
ship with L. peninsularis would scarcely be 
suspected in view of the fact that the latter is 
placed by Pittier in his series Impressinervi 
and that his description includes no reference 
to the conspicuously punctate character of the 
leaflets mentioned by Donnell Smith in his 
original description of the species as Derris 
peninsularis (Bot. Gaz. 44: 111. 1907). Ac- 
tually, Lonchocarpus peninsularis and the 
likewise misplaced L. nicoyensis (Donn. Smith) 
Pittier and L. costaricensis (Donn. Smith) 
Pittier belong to the series Planinervi. Both L. 
peninsularis and L. nicoyensis fall into the 
section Punctati, the former being apparently 
most closely related to L. longistylis Pittier. It 
is doubtless due to the anomalous position in 
Pittier's classification of the frequently col- 
lected L. peninsularis that still another syn- 
onym of this species was created — L. purpusii 

Lonchocarpus monospermus Standi., Field 
Mus. Publ. Bot. 4: 311. 1929 = L. luteomacu- 
latus Pittier, Contr. U. S. Nat. Herb. 20: 64. 

The distinguishing feature ("small one- 
seeded pods") attributed to L. monospermus 
is quite prevalent in L. luteomaculatus, the 
pods of the type specimen of L. luteomaculatus 
(Pittier 4170, U. S. Nat. Herb.) being pre- 
dominently 1-seeded and 2.5 cm long. The 
type of L. monospermus (Standley 53715, 
Chicago Nat. Hist. Mus.) appears to differ in 
no tangible respect from this. 

Lonchocarpus purpusii Brandegee, Univ. 
California Publ. Bot. 6: 500. 1919 = L. penin- 
sularis (Donn. Smith) Pittier, Contr. U. S. 
Nat. Herb. 20: 56. 1917. 

The type specimen of L. purpusii (Purpus 
7849, Univ. Calif.) is a very close match with 
that of Derris peninsularis in the U. S. Na- 
tional Herbarium. For discussion see L. dume- 
torum and L. kerberi above. 

Lonchocarpus seleri Harms, Fedde Rep. 
Spec. Nov. 17: 324. 1921 =L. hondurensis 
Benth., Journ. Linn. Soc. 4: Suppl. 91. 1860. 

L. seleri was set off from L. hondurensis by 
Harms "durch fast sitzende Bliiten auf lan- 
gerem gemeinsamen Stiele und wohl auch 
grossere Vorblatten." Examination of a large 
series of L. hondurensis, however, shows the 
relative length of peduncle and pedicel to be 
very unstable; for example, Mell 530 (U. S.) 
is L. seleri in its peduncles but not in its pedi- 
cels whereas Wilson 706 (U. S.) is L. seleri in 
its pedicels but not in its peduncles. No dif- 
ference was found between the bracts in the 
type fragment of L. seleri in the Chicago 
Natural History Museum {Seler 5052) and 
those of L. hondurensis, which are exceedingly 

Lonchocarpus xuul Lundell, Bull. Torrey 
Bot. Club 69: 391. 1942. 

The reference of this species to the synon- 
ymy of L. guatemalensis Benth. in the recently 
published Flora of Guatemala (Fieldiana, 
Botany, 24(5): 278-279. 1946) seems to be 
clearly erroneous. The long stipes (averaging 
1 cm) of the short, thick pods (generally 2.5-4 
cm long) of L. xuul and its much smaller 
flowers set it off at a glance from L. guatema- 
lensis with its sessile or subsessile, long, flat 
legumes (averaging 6-20 cm long). The author 
of L. xuul correctly indicated its close rela- 
tionship with L. constrictus Pittier among the 
Middle American species. Its nearest ally in the 
genus as a whole is the Venezuelan L. miran- 
dinus Pittier, with which it shows a striking 
similarity in its fruit but from which it differs 
in its few (5 to 9 rather than averaging 15), 
blunt leaflets and in its mainly green flowers. 

The disposition of the following 17 bino- 
mials, the remainder of those proposed from 
Middle America since 1917, must be postponed 
either because no specimens have yet been 
procurable for study or because the material 
so far available has not been sufficient for more 
than tentative conclusions: 

L. apricus Lundell, Lloydia 2: 90. 1939. Chiapas, 

L. belizensis Lundell, Wrightia 1 : 55. British Hon- 

L. castilloi Standi., Tropical Woods 32: 15. 1932. 
Guatemala; British Honduras. 

L. chiapensis Lundell, Wrightia 1: 152. 1946. 
Chiapas, Mexico. 

L. cruentus Lundell, Wrightia 1: 55. 1945. To- 
basco, Mexico. 

L. fuscopurpureus Brandegee, Univ. California 
Publ. Bot. 10: 405. 1924. Veracruz, Mexico. 



L. galleotianus Harms, Fedde Rep. Spec. Nov. 17: 

322. 1921. Oaxaca, Mexico. 
L. gillyi Lundell, Wrightia 1: 56. 1945. Chiapas, 

L. hidalgensis Lundell, Wrightia 1: 153. 1946. 

Hidalgo, Mexico. 
L. hintoni Sand with, Kew Bull. Misc. Inf. 1936: 

4. 1936. Mexico and Guerrero, Mexico. 
L. malacotrichus Harms, Fedde Rep. Spec. Nov. 

17: 323. 1921. Mexico. 
L. monofoliaris Schery, Ann. Missouri Bot. Gard. 

30: 89. 1943. Panama. 

L. nicaraguensis Lundell, Wrightia 1: 154. 1946 

L. phlebophyllus Standi. & Steyerm., Field Mus. 
Publ. Bot. 23 (2): 56. 1944. Guatemala. 

L. stenodon Harms, Fedde Rep. Spec. Nov. 17: 
324. 1921. Oaxaca, Mexico. 

L. whitei Lundell, Wrightia 1: 154. 1946. Nica- 

L. yoroensis Standi., Field Mus. Publ. Bot. 9 (4) : 
296. 1940. Honduras. 

ZOOLOGY. — On the crayfishes of the Limosus section of the genus Orconectes 
(Decapoda: Astacidae)} Horton H. Hobbs, Jr., Miller School of Biology, 
University of Virginia. (Communicated by Fenner A. Chace, Jr.) 

In the course of working over a series of 
crayfishes from the Nashville, Tenn., area 
collected for me by Dr. C. S. Shoup, of 
Vanderbilt University, and Dr. Mike 
Wright, of Tusculum College, it was neces- 
sary to examine the type specimens of 
several of the species of the Limosus section. 
In making this study I arrived at certain 
conclusions, which are discussed below, 
concerning the affinities of the members of 
this section which are not in accord with the 
opinions of others. In addition to the de- 
scription of a new species I am including a 
key to the species of the Limosus section. 
The new species herein described was 
first reported by Fleming (1939) under the 
name Cambarus propinquus sanborni Faxon. 
I have compared my specimens with Flem- 
ing's description and figures which leave 
much to be desired. In addition, I have 
examined several crayfish he sent to the 
United States National Museum 2 from the 
only locality he cited for his C. propinquus 
sanborni, and I am convinced that his 
specimens were members of the species 
I am . describing below. The only locality 
that Fleming recorded is "Mill creek, lo- 
cated about 4 miles south of Nashville 
where this creek crosses the Murfreesboro 
Road [U. S. Hy. 41], studied on August 18, 
1935" (Fleming, 1939, 13: 298) . 3 

1 Received July 17, 1947. 

2 These are not the specimens mentioned by 
Fleming in his report of the higher Crustacea in 
the Nashville region, for he stated that Mill 
Creek was "studied on August 18, 1935," and 
these were collected on August 15, 1936. 

3 This peculiar citation is necessary because 

Genus Orconectes Cope 1872 
Orconectes shoupi, n. sp. 4 

Cambarus propinkuus Fleming, 1939, 14: 305 

(in part). 
Cambarus propinquus sanborni Fleming (not 

Faxon), 1939, 14: 305, 306 (in part), 319, 320, 

and pi. 14. 

Diagnosis. — Rostrum with lateral spines, 
margins thickened and concave laterad; upper 
surface with or without a median carina. Fin- 
gers of chela with usual longitudinal ridges much 
reduced; whole hand resembling that of 0. 
rusticus placidus (Hagen, 1870: 65). Epistome 
with a median carina (see Fig. 4). Areola ap- 
proximately 9 to 10 times longer than broad, 
with two or three punctations in narrowest 
part — length 34-36 percent of entire length of 
carapace; in male, hooks on ischiopodites of 
third pereiopods only. Terminal elements of 
first pleopod of first-form male short, reaching 
almost to coxopodite of second pereiopod. Two 
terminal elements separated for only a short 
distance near tip: mesial process recurved 
caudomesiad and shorter than central pro- 
jection. Annulus ventralis immovable. (See 
Fig. 5 for surface contour.) 

Holotypic male, form I. — Bod}' subovate, 

Fleming's paper was divided, and appeared in 
two volumes- of the Proceedings of the Tennessee 
Academy of Sciences; an overlapping in page 
references causes a further complication. See 
"Literature Cited." 

4 Dr. C. S. Shoup has made a definite and 
worth-while contribution toward a knowledge of 
the fauna of the State of Tennessee. In token of 
the interest he has shown in my work on the cray- 
fishes and the many specimens he has added to 
my collection, I name this new species in his 

Jan. 15, 1948 hobbs: crayfishes of genus orconectes (limosus section) 


distinctly depressed. Abdomen narrower than 
thorax. Width of carapace greater than depth 
in region of caudodorsal margin of cervical 
groove (15.2-9.4 mm). 

Areola moderately narrow (9.6 times longer 
than broad), with two or three punctations 
in narrowest part; cephalic section of carapace 
about 1.8 times as long as areola (length of 
areola about 35.6 percent of entire length of 

Rostrum with thickened margins concave 
laterad. Upper surface concave, but bearing a 
weak median carina. Base of acumen set off by 
corneous knoblike tubercles directed cephalo- 
dorsad. Acumen long and terminating cephalad 
in a corneous knob similarly disposed as the 
tubercles at its base. Subrostral ridges promi- 
nent and visible in dorsal aspect to base of 
acumen. Raised lateral margins of rostrum 
flanked laterally and mesially by rows of 
prominent setiferous punctations. 

Postorbital ridges prominent, grooved dorsad 
and terminating cephalad in heavy acute tu- 
bercles. Suborbital angle absent. Branchio- 
stegal spine obtuse, very much reduced. Small 
lateral spine present on each side of carapace. 
Surface of carapace granulate laterally and 
bearing prominent punctations dorsally; small 
polished area in gastric region. 

Cephalic section of telson with two spines 
in each caudolateral corner. 

Epistome bell-shaped in profile with a me- 
dian longitudinal ridge; no cephalomedian pro- 

Antennules of the usual form, with a small 
spine present on ventral surface of basal seg- 

Antennae broken in holotype but extending 
caudad to cephalic margin of telson in other 
specimens. Antennal scale of moderate width 
with subparallel mesial and lateral margins; 
outer portion broad and swollen and terminat- 
ing distad in a heavy spine; lamellar portion 
broad (see Fig. 9). 

Chela somewhat depressed; palm inflated; 
prominent setiferous punctations present over 
most of chela. Inner margin of palm with three 
rows of squamous ciliated tubercles. Fingers 
widely gaping at base. Upper surface of im- 
movable finger with a narrow well-defined 
ridge along mesial margin; lateral and lower 
margins with prominent punctations; upper 
opposable margin with a row of 18 rounded 

corneous tubercles; an additional prominent 
tubercle present below this row at base of distal 
fifth of finger; minute denticles occurring in a 
single row on penultimate fifth of mesial sur- 
face of immovable finger; mesial distal fifth 
with a broader zone of similar denticles; lower 
proximomesial surface bearded. Opposable 
margin of dactyl with 23 rounded corneous 
tubercles; distal half of mesial margin bearing 
minute denticles interspersed between the 
rounded tubercles. Otherwise dactyl similar to 
immovable finger. 

Carpus of first pereiopod longer than broad, 
with a prominent longitudinal furrow on 
upper surface; all surfaces with scattered punc- 
tations. Mesial surface with a heavy spinous 
tubercle; distal upper mesial margin with a 
prominent rounded tubercle; lower distal mar- 
gin with two heavy tubercles. 

Merus, viewed laterally, with a single promi- 
nent tubercle on upper distal surface (a some- 
what less prominent one lying mesiad of it but 
not evident in lateral aspect, nor is it present 
on sinistral merus). Lateral and mesial surfaces 
sparsely punctate. Lower surface with a lateral 
row of five small tubercles and a mesial row of 
eight (only the distal one in each row at all 

Hooks on ischiopodites of third pereiopods 
only ; hooks strong with proximal surfaces sub- 
plane and bearing setae. 

First pleopod almost reaching coxopodite of 
second pereiopod when abdomen is flexed. Tip 
terminating in two distinct parts, which are 
separated for only a short distance. Central 
projection corneous, almost straight, and some- 
what bladelike, with tip slightly recurved. 
Mesial process extending distad for the proxi- 
mal half of its length, then bending somewhat 
sharply caudomesiad. 

Morphotypic male, form II. — The only sec- 
ond-form male collected from the type locality 
is immature. Most of the tubercles mentioned 
in the description of the first-form male are 
present in this specimen as acute spines. The 
lower surface of the carpus and the cephalo- 
mesial surface of the merus of the cheliped with 
tufts of long plumose setae. Rostrum without 
median carina. Hook on ischiopodite of third 
pereiopod very much reduced. See Figs. 16 and 
19 for structure of first pleopod of a mature 
second-form male from Mill Creek. 

Allotypic female. — Differs from the holo- 



Figs. 1-14. — 1, Dorsal view of carapace of Orconectes shoupi, n. sp.; 2, lateral view of same; 3, upper 
surface of chela of first-form male, 0. shoupi; 4, epistome of 0. shoupi; 5, annulus ventralis of 0. shoupi; 
6, mesial view of first pleopod of first-form male of 0. pellucidus australis (Rhoades), from McFarlen 
Cave, SW£ NW1 sec. 22, T. 3, R. 3 E., near Garth, Jackson County, Ala.; 7, mesial view of first pleopod 
of first-form male of 0. pellucidus pelluicdus (Tellkampf), from Mammoth Cave, Roaring River, Ed- 
monson County, Ky.; 8, caudal view of first pleopods of first-form male, 0. pellucidus packardi Rhoades 
(holotype), from Cumberland Crystal Cave at Alpine, Pulaski County, Ky.; 9, antennal scale of 0. 
shoupi; 10, mesial view of first pleopod of first-form male, of 0. inermis Cope, from Seibert's Well Cave 
near Wyandotte Cave, Crawford County, Ind.; 11, mesial view of first pleopod of first-form male of 
0. pellucidus packardi Rhoades (holotype) (see explanation of Fig. 8); 12, caudal view of first pleopods 
of first-form male of 0. pellucidus pellucidus (Tellkampf) (see explanation of Fig. 7); 13, caudal view 
of first pleopods of first-form male of 0. pellucidus australis (Rhoades) (see explanation of Fig. 6); 
14, caudal view of first pleopods of first-form male of 0. inermis Cope (see explanation of Fig. 10). 

Jan. 15, 1948 hobbs: crayfishes of genus orconectes (limosus section) 


Figs. 15-28 (all figures except 15, 16, and 19 lateral views of the first pleopods of first-form males). — 
15, Mesial view of first pleopod of first-form male of Orconectes shoupi, n. sp.; 16, same, second-form 
male; 17, 0. harrisoni (Faxon), from stream at Irondale, Washington County, Mo.; 18, 0. sloani 
(Bundy), from Little Creek, Jefferson Township, Preble County, Ohio; 19, lateral view of first pleopod 
of second-form male of 0. shoupi; 20, 0. shoupi; 21, 0. rafinesquei Rhoades (holotype), from Rough 
River, at Falls-of-Rough, Grayson- Breckinridge Counties, Ky.; 22, 0. limosus (Rafinesque), from Le- 
man Place, Lancaster County, Pa.; 23, 0. tricuspis Rhoades (holotype), from Pete Light's Spring, 
3 miles east of Canton, Trigg County, Ky. ; 24, 0. indianensis (Hay), no locality given, U.S.N.M. no. 
44448; 25, 0. propinquus propinquus (Girard), from Rocky Creek, Muncie County, 111.; 26, O. difficilis 
(Faxon),|from stream 1 mile south of Wilburton, Latimer County, Okla.; 27, 0. kentuckiensis Rhoades 
(holotype), from Piney Creek, 3 miles west of Shady Grove, Crittenden County, Ky.; 28, 0. sloani 
(Bundy), Indiana (probably from near New Albany), U.S.N.M. no. 58058. 



lypic male in that the tubercles are for the 
most part more spiniform; epistome with a 
small cephalomedian spine; upper distal surface 
of merus of cheliped with two prominent 
tubercles evident in lateral aspect; extreme 
distal margin of merus emarginate; lower sur- 
face of merus and mesial surface of carpus with 
tufts of plumose setae. Annulus ventralis sub- 
spindle-shaped, with the greatest length in the 
transverse axis; cephalic margin evenly rounded 
and firmly fused with sternum; sinus originates 
near cephalomedian margin, extends caudad 
for a short distance, and turns gently caudodex- 
trad, then abruptly sinistrad to cross the me- 
dian line; here it turns caudad and slightly dex- 
trad to the median line and then caudad to the 
midcaudal margin of the annulus (see Fig. 5). 

Measurements. — Holotypic male : Cara- 
pace, height 9.4, width 15.2, length 26.9 mm; 
areola, width 1.0, length 9.6 mm; rostrum, 
width 3.8, length 6.4 mm; abdomen, length 
27.7 mm; right chela, length of inner margin of 
palm 7.5, width of palm 11.4, length of outer 
margin of hand 28.7, length of dactyl 19.3 mm. 
Allotypic female: Carapace, height 8.0, 
width 11.6, length 22.6 mm; areola, width 
0.70, length 7.8 mm; rostrum, width 3.2, 
length 5.9 mm; abdomen, length approx. 24 
mm; right chela, length of inner margin of 
palm 5.2, width of palm 7.3, length of outer 
margin of hand 17.2, length of dactyl 11.3 mm. 

Type locality. — Mill Creek, tributary of 
Cumberland River, east of Oglesby near An- 
tioch Pike, 10 miles south of Nashville, David- 
son County, Tenn. Dr. Shoup has kindly fur- 
nished the following information: This creek is 
a hard-water stream flowing over sand and 
rubble and in its upper reaches over limestone 
ledges. The banks are silty and muddy, and 
shade is provided by reeds and trees along its 
banks. Much of its course is through pasture 
and cultivated lands. In riffle areas the water 
has a slightly greenish cast on cloudy days. 
(M. 0. alkalinity — 154.0 p. p.m. on January 
24, 1947.) 

Disposition of types. — The holotypic male, 
the allotypic female, and the morphotypic 
male, form II, are deposited in the United 
States National Museum (no. 84072), and in 
addition five second-form males and one fe- 
male, collected by R. S. Fleming (U.S.N.M. 
no. 77908) are designated as paratypes. Of the 
remaining paratypes, one male, form I, and one 

female are deposited in the University of 
Michigan Museum of Zoology; one male, form 

I, and one female in the Museum of Compara- 
tive Zoology; and 1 1 males, form I, one male, 
form II, two females, five immature males, and 
one immature female are in my personal col- 
lection at the University of Virginia. 

Specimens examined. — Tennessee, David- 
son County: Seven Mile Creek, 5 miles south- 
east of Nashville, November 11, 1944, two 
males, form I, one male, form II, and two 
females — C. S. Shoup, collector; Mill Creek, 

10 miles south of Nashville, November 11, 
1944, eight males, form I, one male, form II, 
and three females — C. S. Shoup, collector; 
Mill Creek at junction with U. S. Highway 41, 
about 3 or 4 miles south of Nashville, October 

II, 1939, two males, form I— W. K. Smith, 
collector; same locality, August 15, 1936, five 
males, form II, and one female — R. S. Fleming, 
collector; Mill Creek at Antioch Pike, July 
19, 1945, two males, form I, one male, form 

11 (shed test), and four immature males — 
Mike Wright, collector. 

Fleming (1939, 14: 319) states: "All of these 
species (including C. propinquus sanborni 
Faxon) were present throughout the region 
studied . . . "; however, he cites only one lo- 
cality in which this species was taken. I 
strongly doubt that his statement is correct, 
for Drs. Shoup and Wright have collected in a 
large number of localities in the Nashville 
region and have taken 0. shoupi ( = Fleming's 
C. propinquus sanborni Faxon) from only the 
localities cited above. 

Variation. — The rostrum may or may not 
bear a median carina. The bearded condition 
of the cheliped which is pointed out in the 
description of the morphotypic male, form II, 
is best developed in young specimens and may 
be reduced or obsolete in older ones. As in most 
species the spiniform condition is accentuated 
in the younger specimens, and in the older 
ones very much reduced; further, in some of the 
females mirrored images of the annulus ven- 
tralis as described for the allotype occur. 

Relationships. — Orconectes shoupi is a mem- 
ber of the Limosus section; it possesses short 
gonopods, the tips of which are separated for 
only a short distance. Its closest affinities are 
with 0. sloani (Bundy) (1876:24), O. tricus- 
pis Rhoades (1944:117), and 0. rafinesquei 
Rhoades (1944:116). 0. shoupi may readily be 

Jan. 15, 1948 hobbs: crayfishes of genus orconectes (limosus section) 


distinguished from any other species of the 
Limosus section by the rostrum with thickened 
ridges and the long-fingered chelae — both of 
which resemble those of 0. rusticus placidus. 
(see further remarks below.) 

Limosus Section 

Ortmann (1931:64) defined the section of 
Orconectes limosus as follows: "Gonopods of 
male, short, rather thick up to near the tips, 
reaching to the coxopodites of the third peraeo- 
pod. Tips separated for a short distance only, 
each tapering to a point. Males with hooks on 
third, or on third and fourth peraeopods." In 
this section he included 0. harrisoni, 0. sloani, 
0. indianensis, 0. limosus, 0. pellucidus pel- 
lucidus pellucidus, and 0. pellucidus testii. 

Since 1931 Rhoades has described several 
additional species and subspecies belonging to 
the Limosus section, and in his Crayfishes of 
Kentucky (1944:117) recognized two groups 
of the section, and listed under them the spe- 
cies indicated below: 

Limosus group — "characterized by strongly di- 
verging tips of the gonopods." Species: 0. limosus 
(Rafinesque), 0. sloani (Bundy), and 0. indian- 
ensis (Hay). 

Rafinesquei group — "the tips of the first pleo- 
pod are both recurved in the same direction." 
Species: 0. rafinesquei Rhoades, 0. tricuspis 
Rhoades, 0. pellucidus pellucidus (Tellkampf), 
0. pellucidus testii (Hay), 0. pellucidus australis 
(Rhoades), 0. pellucidus packardi Rhoades, 0. 
kentuckiensis Rhoades, and 0. harrisoni (Faxon) 

It is questionable that the above subdivision 
of the section into the Limosus and Rafinesquei 
groups is based on true affinities: e.g., if the 
pleopod of 0. kentuckiensis is compared with 
that of 0. sloani and 0. tricuspis, certainly it is 
more like that of the former. This relationship 
is seen not only in the first pleopod but also in 
the annuli ventralis of the two. It also seems 
to me that 0. harrisoni is more nearly related 
to 0. sloani than it is to 0. tricuspis or 0. 
rafinesquei. Except for the fact that the ter- 
minal elements of the first pleopods of the 
several subspecies of 0. pellucidus are "re- 
curved in the same direction" (and I might 
indicate that among the specimens I have 
examined of pellucidus pellucidus the terminal 
elements are straight), I can see no indication 
of closer affinities of these forms with the mem- 
bers of the Rafinesquei group than with those 
of the Limosus group — in fact, if any division 

of the Limosus section is made then it would 
seem that 0. inermis and the various sub- 
species of 0. pellucidus would constitute a 
natural group that should receive a status 
equivalent to that of the other subdivisions. 

The problem of the status of 0. inermis re- 
mains unsolved. Though I have seen relatively 
few specimens of the several subspecies of 0. 
pellucidus I have examined several belonging 
to all of them, and none are like 0. inermis. 
Perhaps it will be shown to be a subspecies of 
0. pellucidus; however, until future work will 
indicate intergradation between the two, it 
seems advisable to retain its specific status. 

As might be expected, with the discovery of 
additional species the Limosus section has 
become decidedly less clear cut, and certain 
species exhibit characters transitional between 
the more typical members of the Limosus 
section and members of other sections of the 
genus. Even in Ortmann's diagnosis of the 
Limosus section quoted above the best charac- 
ter is stated on a relative basis, and a worker 
not already familiar with an over-all picture of 
the genus would have difficulty in deciding 
whether a given specimen belonged to the 
Limosus or Propinquus sections (Ortmann, 
1931: 64, 65). The difficulty at the time that 
Ortmann diagnosed the section was not so 
great as it has been since the somewhat 
"atypical" 0. tricuspis, 0. rafinesquei, and 0. 
shoupi have been added to the list of described 
species belonging to the section. In these spe- 
cies the terminal elements of the first pleopod 
are almost as slender and long as are those of 
some of the members of the Propinquus section 
(see Figs. 21, 23, 25). These obvious resem- 
blances as well as the similarities of the annuli 
ventralis and other anatomical features be- 
tween 0. propinquus propinquus (Girard, 
1852: 88) and 0. tricuspis make the distinction 
between the two sections seem somewhat 
unnatural — i.e., there seem to me to be about 
as many resemblances between 0. tricuspis and 
the subspecies of 0. propinquus as between 
0. tricuspis and 0. sloani, 0. limosus, and 0. 
indianensis. Furthermore, considering the 
pleopods alone, 0. kentuckiensis is transitional 
between 0. limosus and 0. sloani on one side 
and 0. difiicilis (Faxon, 1898: 656) on the 
other — the latter at present being relegated to 
the Virilis section (Ortmann, 1931: 90). 
Rhoades (1944: 123) states in reference to the 



VOL. 38, NO. 1 

affinities of 0. kentuckiensis that it "represents 
a more advanced stage in the series of the 
'Group rafinesquei'. The tips are stouter and 
more differentiated and the annulus is more 
depressed as in sloani and other members of the 
'Group limosus'. In this character it resembles 
closely 0. immunis immunis. Furthermore, it 
is not difficult to see a possible affinitj' to the 
'Section of C. virilis' [ = 0. virilis] even in the 
gonopods of the male." 

Whether these similarities are results of 
convergence or whether they indicate actual 
close relationships can hardly be determined 
until a more exhaustive study of the group 
(which will necessarily mean extensive collect- 
ing) is made. On the basis of the evidence at 
hand I find it difficult to consider these re- 
semblances arising independently in the three 

The taxonomists working with the cray- 
fishes of the Cambarinae have for a long time 
found it convenient to recognize "sections," 
"groups," and "subgroups," and even though 
there are certain species that on the bases of the 
diagnostic characters appear to be intermedi- 
ate between two sections or groups, at least a 
temporary retention of their usage seems de- 
sirable. Whereas the limits of variation in the 
three sections of the subgenus Orconectes are 
not decidedly marked, and almost impossible 
to define in words, recourse to determined 
specimens or figures should alleviate difficulty 
in determining to which section or group any 
specimen in question belongs. For this reason 
I am including a sketch of the pleopods of 
all the species and subspecies (except 0. pel- 
lucidus testii, of which I do not have a first form 
male) belonging to the Limosus section, and in 
addition, for comparative purposes, the pleo- 
pods of 0. propinquus propinquus and 0. 


(Based on the First-Form Male) 

1. Body pigmented, eyes well developed 2 

Body not pigmented, eyes reduced 9 

2. Terminal elements of first pleopod subequal 

in length and distinctly divergent (central 
projection directed cephalodistad and me- 
sial process caudodistad) 3 

Terminal elements of first pleopod subequal 
or not subequal in length, but central pro- 
jection never bent cephalodistad — either 
straight, directed caudad, or caudodistad. .4 

3. Lateral surface of carapace with only one 

spine 0. indianensis (Hay) 

Lateral surface of carapace with more than 
one spine 0. limosus (Rafinesque) 

4. Central projection bent caudad at an angle 

greater than 45° . . . . O. harrisoni (Faxon) 

Central projection directed distad or bent 

caudad at an angle less than 45° 5 

5. Terminal elements of first pleopod subequal 

in length or mesial process slightly longer 

than central projection. . . 

0. tricuspis Rhoades 

Mesial process never extending quite so far 
distad as central projection 6 

6. Central projection recurved (caudodistad) 

throughout its length; no median carina 
on rostrum .... 0. kentuckiensis Rhoades 
Central projection not recurved caudodistad 
throughout its length; median carina on 
rostrum present or absent 7 

7. Margins of rostrum thickened and concave 

laterad 0. shoupi Hobbs 

Margins of rostrum not thickened, and sub- 
parallel or convergent up to base of lateral 
spines 8 

8. Terminal elements of first pleopod widely 

separated and thick (heavy); tip of mesial 
process caudomesiad of central projection 

0. sloani (Bundy) 

Terminal elements of first pleopod not widely 
separated, and slender and tapering; tip of 
mesial process caudolaterad of central 
projection 0. rafinesguei Rhoades 

9. Margins of rostrum uninterrupted; acumen 

not distinctly set off from rest of rostrum 

0. pellucidus testii (Hay) 

Margins of rostrum interrupted; acumen dis- 
tinctly set off from rest of rostrum 10 

10. Cephalic margin of pleopod without a shoul- 

der at base of central projection; however, 

either straight or curved 11 

Cephalic margin of pleopod with an angular 
or rounded shoulder 12 

11. Cephalic surface of first pleopod in region of 

central projection straight; mesial process 
directed distad and extending distad be- 
yond central projection 

. . .0. pellucidus pellucidus (Tellkampf) 
Cephalic surface of first pleopod in region of 
central projection curved; mesial process 
directed caudodistad and somewhat lat- 
erad, and not extending distad beyond 
central projection 0. inermis Cope 

12. Shoulder on cephalic margin at base of central 

projection rounded; hooks present only on 

ischiopodites of third pereiopods 

0. pellucidus australis (Rhoades) 5 

5 The holotype of 0. pellucidus australis has a 
small short acute spine (probably corresponding 
to the caudal process seen in many members of 
the genus Procambarus), which in lateral aspect 
lies between the central projection and the mesial 

Jan. 15, 1948 



Shoulder on cephalic margin at base of cen- 
tral projection distinctly angular; hooks 
present on ischiopodites of third and fourth 
pereiopods. 0. pellucidus packardi Rhoades. 


Bundy, W. F. List of Illinois Crustacea, with 
descriptions of new species. Bull. Illinois 
Mus. Nat. Hist. 1: 3-25. 1876. 

Faxon, W. Observations on the Astacidae in 
the United States National Museum and in 
the Museum of Comparative Zoology with 
descriptions of new species. Proc. U. S. 
Nat. Mus. 20: 643-694, 9 pis. 1898. 

Fleming, R. S. The larger Crustacea of the 
Nashville region. Journ. Tennessee Acad. 
Sci. 13(4): 296-324; 14(2): 261-264; 
14(3): 299-324. 1938-39. 

Girard, C. A revision of the North American 
Astaci, with observations on their habits and 

geographical distribution. Proc. Acad. Nat. 

Sci. Philadelphia 6: 87-91. 1852. 
Hagen, H. Monograph of the North American 

Astacidae. Illus. Cat. Mus. Comp. Zool. 

no. 3: 1-109, 11 pis. 1870. 
Hay, W. P. The crawfishes of the State of 

Indiana. 20th Ann. Rep. Indiana Geol. 

and Nat. Res. Surv.: 446-507, 15 figs. 

Ortmann, A. E. Crayfishes of the Southern 

Appalachians and the Cumberland Plateau. 

' Ann. Carnegie Mus. 20(2) : 61-160. 1931. 

Rafinesque, C. S. Synopsis of four new 

genera and ten new species of Crustacea 

found in the United States. Amer. Monthly 

Mag. and Crit. Rev. 2. art. 7(9): 40-43. 

Rhoades, R. The crayfishes of Kentucky, 

with notes on variation, distribution and 

descriptions of new species and subspecies. 

Amer. Midi. Nat. 31 (1): 111-149, 10 figs. 

10 maps. 1944. 

ORNITHOLOGY — The races of the black-throated sunbird, Aethopyga saturata 
(Hodgson). 1 H. G. Deignan, U. S. National Museum. 

For more than 30 years the race of 
Aethopyga saturata common on the moun- 
tains of northwestern Siam has, without any 
direct comparison of specimens, been re- 
corded as sanguinipectus (a name originally 
applied to the form of the Karen Hills). 
Topotypes of sanguinipectus and of its pre- 
sumed synonym waldeni (described from 
Mount Muleyit in Tenasserim) are nowhere 
available in America, but reference to the 
first descriptions and especially to Shelley's 
Monograph of the Nectariniidae, pt. 6, 1878, 
pp. 37-38 and colored plate (where the 
description is taken from the types of 
sanguinipectus and the illustration from the 
types of waldeni), has shown that sanguini- 
pectus has the entire throat (except only 
the center of the chin) metallic blue or 
violet and is thus quite different from the 
Siamese bird, as well as from the several 
populations of Indochine that have been 
masquerading under its name. This dis- 
covery has made necessary a revision of the 
species, with the result that the number of 
races has been increased from the five ac- 
cepted by Delacour (Zoologica 29: 34. 
1944) to nine, of which three are here de- 
scribed for the first time. 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received June 10, 

Material essential to the prosecution of 
this study has been courteously sent me by 
the Museum of Comparative Zoology 
(M.C.Z.), the American Museum of Natu- 
ral History (A.M.N.H.), the Princeton 
Museum of Zoology (P.M.Z.), the Acad- 
emy of Natural Sciences of Philadelphia 
(A.N.S.P.), and the Chicago Natural His- 
tory Museum (C.N.H.M.). 

1. Aethopyga saturata saturata (Hodgson) 

[Cinnyris] Saturata Hodgson, India Rev. and 
Journ. Foreign Sci. and Arts 1 (7): 273. Oct. 
1836 (Nepal). 

Nectarinia hodgsonis [sic] Jardine, Naturalist's 
Library 36 [Nectariniidae]: 240, 269 [where 
spelled hodgsoni], pi. 28, 1843 (Nepal). 

Range. Himalayas, from Garhwal to Bhutan. 

Remarks. The reference to Hodgson's name 
is incorrectly cited by Stuart Baker (Fauna of 
British India, birds, ed. 2, 7: 285. 1930) as 
"Ind. Review, vol. ii, p. 273, 1837." Sherborn 
(Index Animalium, p. 5753) gives "India Rev. 
I. 1837, 273." According to the researches of the 
late C. W. Richmond, the first volume of the 
India Review appeared in 12 monthly install- 
ments from April 1836 to March 1837, and the 
proper citation is that given above. 

2. Aethopyga saturata assamensis 


Cinnyris assamensis McClelland, Quart. Journ. 



Calcutta Med. and Phys. Soc. 1 (3): 322. July 
1837 (Assam). Nomen nudum! 
Cinnyris Assamensis McClelland, Proc. Zool. 
Soc. London 7: 167. Mar. 1840 (Assam; type 
locality here restricted to the neighborhood of 
Sadiya, Sadiya Frontier Tract, Assam Prov- 
ince, India). 

Range. — Assam; Burma (north); Yunnan 

Remarks.— Mayr (Ibis 1938: 302) has shown 
that the bird of northern Burma, while most 
like saturata; differs from it by a broadening 
of the yellow band across the lower back and 
by a slight^ shorter wing and a much shorter 
tail. For this population the old name assamen- 
sis may be revived. 

3. Aethopyga saturata sanguinipectus 

Mthopyga sanguinipectus Walden, Ann. Mag. 
Nat. Hist. (4) 15: 400. June 1875 ("Tonghoo 
hills [Karen-hee]," error; type locality cor- 
rected to "the Tonghoo and Karen-nee hills" 
by Wardlaw Ramsay, in The ornithological 
works of Arthur, Ninth Marquis of Tweeddale, 
p. 414, 1881). 

sEthopyga Waldeni Hume, Stray Feathers 5 (1): 
51. Apr. 1877 (Mount Muleyit, Amherst Dis- 
trict, Tenasserim Division, Burma). 

Range. — Southeastern Burma, from lat. 
19° N. south to lat. 14° N. (but see Remarks). 

Remarks. — Shelley (loc. cit. in introductory 
paragraph above) has observed that "the illus- 
trations of both sexes are taken from the speci- 
mens collected by Mr. Davison, and described 
as M. waldeni by Mr. Hume, who has kindly 
forwarded them to me from India to compare 
with the types of M. sanguinipectus, to which 
species they evidently belong." It may be 
hoped that some latter-day students in London 
will compare the two again to learn whether 
they belong to the same subspecies! 

The one bird of sanguinipectus-type avail- 
able to me is a male from Mount Nwalabo 
(lat. 14° N.), which may or may not be properly 
placed under this name. The synonymy and 
range given above for sanguinipectus must 
therefore be considered only tentative. 

4. Aethopyga saturata anomala Richmond 

JEihopyga anomala Richmond, Proc. U. S. Nat. 
Mus. 22: 319. May 12, 1900 (Khao Sung, pe- 
ninsular Siam at lat. 7°32' N., long. 99°50' E.). 

Range. — Specimens are known from Khao 

Soi Dao (lat. 7°20' N., long. 99°50' E.), Khao 
Nok Ra (lat. 7°25' N., long. 99°55' E.), Khao 
Nam Pliu (lat. 7°35' N., long. 99°50' E.), and 
the type locality, all hills in the range dividing 
the Siamese provinces of Phatthalung and 

5. Aethopyga saturata wrayi Bowdler Sharpe 
JEthopyga wrayi Bowdler Sharpe, Proc. Zool. 
Soc. London 1887, pt. 3: 440, pi. 38, fig. 2. 
Oct. 1 (Larut Range, at elev. 4,400 ft., Perak 
State, Malaya). 

Range. — Mountains of Malaya from north- 
ern Perak to southern Selangor and in Pahang. 

6. Aethopyga saturata petersi, n. subsp. 

Type.— C.N.H.M. no. 76209, adult male, col- 
lected at Pa Kha (lat. 22°32' N., long. 104°18' 
E.), Laokay Province, Tongking, on Decem- 
ber 28, 1929, by J. Delacour and P. Jabouille 
(original number 3312). 

Diagnosis. — The adult male of the new form 
is immediately distinguishable from those of 
A. s. saturata (Nepal), A. s. assamensis (As- 
sam), and A. s. anomala (peninsular Siam) by 
having the lower breast sulphur yellow, con- 
spicuously streaked with blood red. 

From those of A. s. sanguinipectus (Karen 
Hills) and A. s. johnsi (southern Annam) it is 
separable by having the entire area between 
the metallic-blue mustachial streaks, from chin 
to breast, unglossed black (not metallic blue). 

It is, in fact, nearest the geographically re- 
mote A. s. wrayi (Malaya), from which it may 
be known by its having the sulphur-yellow 
breast patch more clearly defined and poste- 
riorly more extensive, the red streaks on the 
breast patch more numerous, and the remain- 
ing under parts a slightly paler, more yellowish, 
less grayish, olive green. 

Range. — Yunnan (southeast); Tongking; 
Annam (north); Haut-Laos; Southern Shan 
States (Kengtung State); northern Siam (Chi- 
ang Rai and Nan Provinces). 

Specimens examined. — 34 males, 3 females. 

Remarks.— This race is named in honor of 
James Lee Peters, of the Museum of Compara- 
tive Zoology. 

7. Aethopyga saturata galenae, n. subsp. 

Type.— U.S.N.M. no. 331074, adult male, 
collected on Doi Langka (Khao Pha Cho), 
northern Siam at lat. 19°00' N., long. 99°25 / E., 

Jan. 15, 1948 

proceedings: the academy 


on November 10, 1930, by Hugh M. Smith 
(original number 4307). 

Diagnosis. — The adult male is nearest that 
of A. s. 'peter si (Tongking) but differs by having 
the sulphur-yellow breast patch posteriorly 
more extensive and less clearly defined from 
the remaining underparts, which are oliva- 
ceous-yellow {not grayish olive-green). 

The adult female is separable from that of 
A. s. peter si by having the underparts suffused 
posteriorly with yellow, rather than a uniform 
grayish olive-green. 

Range. — Siam (northwest). 

Specimens examined. — 18 males, 3 females. 

Remarks. — Of four males from Doi Ang Ka 
(lat. 18°35' N., long. 98°30' E.), three are in- 
separable from topotypical A. s. galenae; the 
exceptional specimen has the majority of the 
feathers of the upper half of the throat wholly 
or partly metallic blue and thus shows ap- 
proach to A. s. sanguinipectus-waldeni of 
Mount Muleyit. It is highly probable that 
birds of the unexplored hills lying between Ang 
Ka and Muleyit will prove to be true sanguini- 

This race is named in honor of my mother. 

8. Aethopyga saturata ochra, n. subsp. 

Type.— C.N.H.M. no. 91613, adult male, 
collected at Thateng (lat. 15°31' N., long. 
106°22' E.), Sara vane Province, Bas-Laos, on 
November 26, 1931, by J. Delacour (original 
number 312). 

Diagnosis. — The adult male is nearest that of 
A. s. galenae (northwestern Siam), from which 
it differs (and so much the more from that of 
A. s. peter si) by having the sulphur-yellow 
breast patch posteriorly still more extensive 
and scarcely defined at all from the remaining 
underparts, which are more strongly suffused 
with yellow then in any other race. 

The adult female seems to be identical with 
that of A. s. galenae. 

Range. — Bas-Laos (Boloven Plateau). 

Specimens examined. — 7 males, 2 females. 

9. Aethopyga saturata johnsi Robinson and 
Boden Kloss 

JEihopyga sanguinipectus johnsi Robinson and 
Boden Kloss, Ibis (11) 1 (4): 621. Oct. 23, 1919 
(Dran, Haut-Donai Province, southern An- 
nam). • 

Range. — Southern Annam (Langbian Pla- 


413th meeting of board of managers 

The 413th meeting of the Board of Man- 
agers, held in the Cosmos Club, December 15, 
1947, was called to order at 8:10 p.m. by the 
President, Dr. Waldo L. Schmitt. Others 
present were: H. S. Rappleye, H. A. Rehder, 
H. B. Collins, Jr., F. G. Brickwedde, F. L. 
Mohler, J. S. Wade, W. W. Rubey, L. E. 
Yocum, W. A. Dayton, M. A. Mason, C. L. 
Garner, and C. L. Gazin. 

Eighteen persons were elected to member- 
ship. Dr. Percy W. Bridgman, a member of 
the Academy and a Nobel Prize winner, who 
has spoken before the Academy, was proposed 
for Honorary Membership. The Board voted 
unanimously that he be awarded this distinc- 
tion. Ten nominations for resident membership 
were presented. 

The Secretary, Dr. C. L. Gazin, reported the 

death of Dr. Dwight F. Windenburg on 
November 14, 1947, formerly with the David 
Taylor Model Basin. He also reported the resig- 
nations of Drs. Albert E. Longley and John 
B. Mertie, Jr. The Board accepted these resig- 
nations, with regret, effective December 31, 
1947. The request of Dr. H. C. Oberholser, 
who has paid dues from 1906 through 1947 
and who has retired from the gainful practice 
of his profession, that he be placed on the re- 
tired list of Academy members was approved 
by the Board. 

The Secretary announced that he received, 
on November 28, 1947, the nomination of Dr. 
Gilbert Grosvenor for President of the 
Academy in 1948. [This candidate subsequently 
withdrew. — Ed.] 

The Treasurer, H. S. Rappleye, requested 
that he be authorized to transfer the present 
two investment certificates, amounting to 
$5,000 in total, in the First Federal Savings & 



Loan Association, to a single account-book 
form as proposed by the Association. This re- 
quest was granted. 

The meeting was adjourned at 9:42 p.m. 
C. L. Gazin, Secretary. 


There follows a list of persons elected to 
membership in the Academy, by vote of its 
Board of Managers, since December 16, 1946, 
who have since qualified as members in ac- 
cordance with the bylaws. (See also previous 
list in May 15, 1947, issue of the Journal.) The 
bases for election are stated with the names of 
the new members. 


Elected December 16, 1946 

Walter A. McCubbin, botanist, Bureau of 
Entomology and Plant Quarantine, in recog- 
nition of his contributions to the scientific basis 
of plant quarantines and the means of distribu- 
tion of parasitic fungi. 

Dean B. Cowie, physicist, Department of 
Terrestrial Magnetism, Carnegie Institution of 
Washington, in recognition of his work in the 
construction of the cyclotron at the Depart- 
ment, for many cooperative researches in which 
cyclotron products made by him were used, and 
especially his work on the distribution of anti- 
mony in Filaria infections. 

Elected January 13, 1947 

Mark W. Woods, botanist, National Cancer 
Institute, in recognition of his studies on the 
cytology of plant virus diseases and the origin 
and nature of viruses. 

John W. Aldrich, biologist, U. S. Fish and 
Wildlife Service, in recognition of his contribu- 
tions to knowledge of the taxonomy and dis- 
tribution of North American birds. 

Elected April 14, 1947 

Philip Drucker, anthropologist, Bureau of 
American Ethnology, Smithsonian Institution, 
in recognition of his contribution to the eth- 
nology of the Northwest coast and to the 
archaeology of these regions and of Veracruz, 

Harriet L. Frush, chemist, National Bu- 
reau of Standards, in recognition of her con- 
tribution to organic chemistry, more especially 

the development of the first satisfactory mech- 
anism for the production of orthoesters from 
transacetohalogen compounds and the de- 
velopment of methods for the separation of 
manuric and galacturonic acids. 

Camille L. Lefebvre, botanist, Bureau of 
Plant Industry, Soils, and Agricultural Engi- 
neering, in recognition of his researches in plant 
pathology, especially fungous diseases of forage 

Curtis W. Sabrosky, entomologist, Bureau 
of Entomology and Plant Quarantine, in recog- 
nition of his contributions to the classification 
of the Diptera and in particular his work on the 
taxonomy of the family Chloropidae. 

Walter B. Lang, geologist, U. S. Geological 
Survey, in recognition of the pioneer study of 
the stratigraphy and sedimentation of the 
Permian of the Southwestern United States 
and its application to the discovery of potash 
and the development of a self-sufficient potash 

Louise M. Russell, entomologist, Bureau 
of Entomology and Plant Quarantine, in recog- 
nition of her contributions to the taxonomy of 
the sternorhynchous Homoptera, in particular 
her studies on the classification of the coccid 
genus Asterolecanium, on various groups of the 
whitefly family Aleyrodidae, and on various 
species of the Psyllidae. 

George M. Foster, Jr., anthropologist, 
Institute of Social Anthropology, Smithsonian 
Institution, in recognition of his contributions 
to the development of the science of ethnology 
in Mexico. 

Fenner A. Chace, Jr., biologist, U. S. Na- 
tional Museum, in recognition of his contribu- 
tions in zoology, especially in the field of 

Ruth E. Gordon, bacteriologist, Bureau of 
Plant Industry, Soils, and Agricultural Engi- 
neering, in recognition of her contributions to 
bacteriology, especially her. researches on the 
acid-fast and on the spore-forming bacteria. 

Elected October 6, 1947 

Hubert R. Snoke, engineer, National Bu- 
reau of Standards, in recognition of his impor- 
tant work on roofing and building materials. 

Elliott B. Roberts, engineer, U. S. Coast 
and Geodetic Survey, in recognition of his con- 
tributions in the development of oceanographic 
instruments and techniques, and in exploration 

Jan. 15, 1948 

proceedings: philosophical society 


and geodetic surveying in the Philippine Is- 
lands, the Aleutian Islands, and other areas. 

Curtis J. Humphreys, physicist, National 
Bureau of Standards, in recognition of his work 
on spectroscopy, including Zeeman effect and 
intensity measurements in the spectrum of the 
solar chromosphere. 

Richard K. Cook, physicist, National Bu- 
reau of Standards, in recognition of his con- 
tributions to acoustics and in particular to the 
absolute measurement of sound intensity. ■ 


Elected October 6, 1947 

Theodore E. White, paleontologist, Mu- 
seum of Comparative Zoology, Cambridge, 
Mass., in recognition of his contributions to 
vertebrate paleontology, in particular to the 
morphology of primitive tetrapods. 

1268th Meeting 

The 1268th meeting, the occasion of the 
Fifteenth Joseph Henry Lecture, was held in 
the Auditorium of the U. S. National Museum, 
May 25, 1946, President Defandorf presiding. 

Program: Harvey Fletcher, Bell Tele- 
phone Laboratories: The pitch, loudness and 
quality of musical tones. This paper has been 
published in full in the American Journal of 
Physics 14: 215-225. 1946. 

1269th Meeting 

The 1269th meeting was held in the Cosmos 
Club Auditorium, October 12, 1946, President 
Defandorf presiding. 

Program: H. W. Wells, Department of 
Terrestrial Magnetism: Panoramic ionospheric 
recorder. — The application of a new technique 
for ionospheric research, developed at the 
Department of Terrestrial Magnetism, Car- 
negie Institution of Washington, under spon- 
sorship of the United States Army Signal 
Corps, has led to the discovery of hitherto un- 
suspected rapid motions and fluctuations. The 
panoramic ionospheric recorder in develop- 
ment, operating over a range 1.0 to 20.0 
megacycles per sec and registering the iono- 
spheric echoes on a single frame of 16-mm 
motion-picture film, was described briefly. This 
apparatus makes two records each minute in 
normal operation. Projection of the series of 

16-mm records thus obtained as a motion pic- 
ture gives a condensation of time scale and a 
continuity of events which reveal ionospheric 
fluctuations of a surprising nature. 

During periods of ionospheric and magnetic 
storms, rapidly moving clouds are "tracked'* 
into the F2 (outer) layer. Upon merging of the 
clouds with the F2 layer the ionization is seen 
to increase, and downward rippling motions 
suggest an extension of the effect into lower 
regions of the ionosphere. 

The technique of recording offers consider- 
able promise both as a research tool leading to 
a better understanding of atmospheric physics 
and as an educational aid in training of per- 
sonnel for ionospheric investigations. (Author's 

Newborn Smith, National Bureau of Stand- 
ards (paper read by Mr. Silberstein) : Longi- 
tude effect in F2-layer characteristics. — Before 
adequate world-wide ionospheric observations 
had been made, it was assumed that the 
monthly average F2-layer critical frequencies 
and virtual heights were the same, at the same 
local time, for stations at the same geographic 
latitude but different longitudes. Radio operat- 
ing data indicated, however, that this assump- 
tion was questionable, and ionospheric observa- 
tions from the world-wide network of stations 
established during the war showed a pronounced 
longitudinal variation, apparently associ- 
ated with differences in geomagnetic latitude. 
This paper presented the evidence for the 
longitude effect, indicating how it led to the 
operational use of separate ionospheric predic- 
tion charts for the different world zones. 
(Author's abstract.) 

1270th Meeting 

The 1270th meeting was held in the audi- 
torium of the Cosmos Club, October 26, 1946, 
President Defandorf presiding. 

Program: N. P. Hays, Wright Field: Polar 
navigation. — Major Hays discussed his part as 
navigator in the preparations and flight of the 
Pacusan Dreamboat from Oahu to Cairo by 
way of the geomagnetic pole. (Secretary's ab- 

Sir Robert Watson- Watt, Radio Board 
of Great Britain: Air navigation by radio meth- 
ods. — The early history, development, and use 
of radio were discussed. The use of radio re- 
flections from the ionosphere and other small 



reflections was also emphasized. (Secretary's 

1271st Meeting 

The 1721st meeting was held in the audi- 
torium of the Cosmos Club, November 9, 1946, 
President Defandorf presiding. 

Program: J. H. Curtiss, National Bureau of 
Standards: High-speed digital computing ma- 
chines, Pt. 1. — The chief purpose of this paper 
was to list and describe the various current 
Federally financed projects in the field of 
automatic digital computing machines. Those 
already in existence, listed in the approximate 
order of completion, are: the "IBM Automatic 
Sequence Controlled Calculator" at Harvard 
University; two small Bell Telephone Labora- 
tories relay computers, of which one is at the 
Naval Research Laboratory and the other is at 
Fort Bliss, Tex.; five small relay calculators 
developed by the International Business Ma- 
chines Corporation, of which one is at the 
Naval Proving grounds at Dahlgren, two are 
at the Ballistics Research Laboratory of the 
Aberdeen Proving Grounds, and two are at 
Watson Laboratories in New York City; and 
finally, the electronic "ENIAC," built at the 
University of Pennsylvania for Aberdeen. Two 
large Bell Telephone relay computers are al- 
most finished; one is for Aberdeen and the other 
for the Langley Memorial Laboratory of 
N.A.C.A. Harvard University is constructing 
a large relay machine for the Naval Proving 

Important development projects are: (1) At 
the National Bureau of Standards, construc- 
tion of large-capacity digital computers for 
the U. S. Navy and the Bureau of the Cen- 
sus, and a component development project 
for the Ordnance Department of the Army; 
(2) at the University of Pennsylvania, con- 
struction of another electronic digital computer 
for Aberdeen; (3) at the Institute for Advanced 
Study and RCA Laboratories at Princeton, 
N. J., construction of an electronic digital 
computer; (4) at the Massachusetts Institute 
of Technology, construction of a large electron- 
ic digital simulator for the Navy and also a 
component research project financed by the 
Rockefeller Foundation. (Author's abstract.) 

S. N. Alexander, National Bureau of 
Standards: High-speed digital computing ma- 
chines, Pt. 2. — There are two basic approaches 

by which scientific computations can be re- 
duced to an automatic operation. These are the 
analogue and the digital procedures. The ana- 
logue procedure had been exploited to the 
greater extent prior to the war. During the 
war the need for greater accuracy and flexi- 
bility led to intensive development of the digi- 
tal procedure. The first of these machines, 
using electromechanical devices, was developed 
jointly by Harvard University and the In- 
ternational Business Machines Corporation. 
Next, the Bell Telephone Laboratories de- 
veloped a computing machine using electrical 
relays. Finally, an essentially electronic ma- 
chine of greatly increased operating speed was 
constructed at the University of Pennsylvania 
for the Army Ordnance. These applications of 
electromechanical and electronic techniques 
pointed the way toward making available the 
flexibility that is inherent in the digital pro- 
cedure. An immediate goal is to apply this 
flexibility to the solution of partial differential 
equations. To accomplish this will require the 
development of still faster electronic machines 
of much greater capacity. Consequently, the 
present program for the development of large- 
scale, high-speed digital computing machines 
is being pursued with even more vigor than 
during the war. (Author's abstract.) 

1272d Meeting 

The 1272d meeting was held in the audi- 
torium of the Cosmos Club, November 23, 
1946, President Defandorf presiding. 

Program: R. J. Seeger, Naval Ordnance 
Laboratory: Shock-wave phenomena. — A signifi- 
cant property of linear differential equations 
is the principle of superposition of solutions, 
e.g., Huygens's principle in physics. Is there a 
corresponding principle in the case of nonlinear 
differential equations? Shock-wave phenomena 
presents a means of studying this question from 
the viewpoint of Nature's own integrations. 
This paper formed essentially a report of the 
research along this line initiated at the Bureau 
of Ordnance (Navy Department) by J. Von 
Neuman. It started with optical illustrations 
of shock waves occurring in ballistics, air jets, 
wind tunnels, and explosions. The mathemati- 
cal concept of a shock wave was presented, 
leading to the simple step-shock model de- 
scribed by the Rankine-Hugoniot equations. 
This model has been used in the investigation 

Jan. 15, 1948 

proceedings: philosophical society 


of the interaction of shock waves. The theoreti- 
cal predictions for reflection from a rigid wall 
were compared with the experimental obser- 
vations. For so-called regular reflection the 
simple model is adequate. In the case of Mach 
reflection, which involves an intersection of 
three shock waves (and a temperature dis- 
continuity) moving away from the wall, there 
is an enigma that can not be solved by any 
simple physical picture such as uniform pres- 
sure between the shocks or even by Prandtl- 
Meyer modification of this uniformity. (Au- 
thor's abstract.) 

F. J. Weyl, Naval Ordnance Laboratory: 
Optical analysis of supersonic flow. — The varia- 
tions in density, which will be found whenever 
a compressible fluid is engaged in nonuniform 
motion, imply accompanying variations in the 
local index of refraction. Consequently a beam 
of transluminating light will be disturbed by 
the presence of flow without, in its turn, dis- 
turbing the flow. This fact is used in three 
basically different fashions, depending on the 
optical apparatus inserted into the light path, 
to obtain records of the density variations in 
the compressible flows. Referred to as the 
shadowgraph, the schlieren, and the inter- 
ferometric methods, they render visible re- 
spectively the total distortion, the change in 
direction, and the change in the time of transit 
experienced by the transluminating light beam 
on account of the presence of the flow. These 
three methods were first discussed from a 
quantitative analytic viewpoint and were then 
illustrated individually by typical records al- 
lowing a direct comparison of their respective 
strengths and weaknesses. (Author's abstract.) 

1273d Meeting 

The 1273d meeting, the 76th annual meeting, 
was held in the auditorium of the Cosmos Club, 
December 7, 1946, President Defandorf pre- 

The report of the treasurer, confirmed by the 
report of the auditing committee, showed an 
income from dues and interest on investments 
of $1357.54 and expenditures of $1137.20, 
leaving a net gain of $220.34 on ordinary ex- 
penses. The ordinary expenses were at the 
rate of $3.27 per member. The total estimated 
assets of the Society showed a decrease of 
$271.76 from $16,694.19 as of December 1, 
1945, to $16,377.43 as of December 1, 1946. 

The joint report of the Secretaries showed an 
active membership of 359 as of December 1, 
1946 — a net increase of 12 members during the 
year. The following persons were elected during 
the year: W. D. Appel, C. E. Barthell, 
R. D. Bennett, T. J. Carroll, G. E. Davis, 
R. F. Geller, P. H. Girouard, D. Harkin, 
W. N. Harrison, Mary W. Hodge, H. J. 
Hoge, R. L. McKinney, D. E. Parsons, 
G. W. Peterson, E. K. Plyler, B. P. Ram- 
say, E. B. Roberts, L. H. Rumbaugh, G. B. 
Sabine, G. H. Shortly, H. S. Stewart, R. D. 
Stiehler, E. J. Stovall, Jr., R. Weller, 
F. J. Weyl, W. J. Wyatt. 

The Secretaries reported the deaths of 
Paul Brockett, W. N. Hamilton, and W. P. 

Following the report of the Committee on 
Elections, these officers were declared elected 
for the year 1947: President, Fred L. Mohler; 
Vice-Presidents, Walter Ramberg and Ken- 
neth L. Sherman; Recording Secretary, 
Walter L. Cheney; Treasurer, Frank C. 
Kracek; M ember s-at-Large of General Com- 
mittee, Karl F. Herzfeld, and Lawrence A. 

Program: R. P. Peterson, Johns Hopkins 
Applied Physics Laboratory: Exploration of the 
upper atmosphere by rockets, Pt. 1. — A review 
of the advancements in cosmic rays shows that 
measurement vs altitude has been important 
in illuminating the subject. The measurement 
of residual ionization vs altitude by Hess and 
by Koehorster in 1912 gave the first positive 
evidence of the extraterrestrial nature of the 
cosmic radiation. Later, the measurements of 
Bowen, Millikan, and Neher of ionization vs 
altitude at various latitudes indicated the 
charged-particlelike character of the primaries 
and gave an energy spectrum for the primary 
radiation. At present there remain important 
unsolved problems; the mesatron component 
increases to the highest altitudes to which meas- 
urements have been made, yet its radioactive 
character insists that it is a secondary and 
therefore that it must eventually decrease; 
there is reported an overabundance of soft 
component at higher altitude ; primary particles 
and their basic interactions (at energies 10,000 
times the highest now sought for in man-made 
machines), which produce the soft and mesa- 
tronic secondaries appear to be largely observ- 
able only above the highest balloon flights. 



This paper described the G. M. telescope of 
the type that Schein has used. This telescope 
measures total ionizing flux and compares the 
penetrating ionizing flux with the penetrating 
nonionizing flux, which converts in 2 cm of 
lead. The paper also described the arrangement 
used in the V-2 and presented a short discussion 
of the results obtained. There was indicated a 
need for additional verification, because of a 
rather large and unexpected presence of a 
nonionizing component of the primaries, or 
energy greater than 10 9 EV and capable of 
converting in 2 cm of lead. 

Finally, there was presented a general de- 
scription of data recovery from such rocket 
flights including results of aerodynamic spoiling 
to obtain lower landing speeds for the recovery 
of mechanical recorders, air sample bottles, 
cameras, spectrographs, etc. (Author's abstract.) 

H. E. Newell, Naval Research Laboratory: 
Exploration of the upper atmosphere by rockets, 
Pt. 2. — The principal features of the V-2 rocket 
and its usefulness as a vehicle for upper-at- 
mosphere experimentation were discussed. 
Following a review of the general aspects of 
upper atmosphere research including a dis- 
cussion of the recovery problems, there was a 
discussion of the Naval Research Laboratory's 
work on solar spectroscopy, cosmic rays, pres- 
sure measurements, and ion-density measure- 
ments in the ionosphere. The paper was illus- 
trated by means of slides pertaining to the 
October 10 V-2 flight at White Sands, N. Mex. 
(Author's abstract.) 

1274th Meeting 

The 1274th meeting was held in the audi- 
torium of the Cosmos Club, January 4, 1947, 
President Mohler presiding. 

The retiring president, Francis M. De- 
fandorf, National Bureau of Standards, de- 
livered his presidential address on the subject 
The measurement of high voltage. This address 
will be published in full in the Journal. 

1275th Meeting 

The 1275th meeting was held in the audi- 
torium of the Cosmos Club, January 18, 1947, 
President Mohler presiding. 

Program: R. S. Burington, Bureau of 
Ordnance: The concept (and misconcept) of 
equivalence.— This paper in expanded form is 

published in this number of the Journal (38: 
1-11. 1948). 

1276th Meeting 

The 1276th meeting was held in the audi- 
torium of the Cosmos Club, February 1, 1947, 
President Mohler presiding. 

Program: James M. Klaasse, Naval Ord- 
nance Laboratory: Airborne magnetometer in 
geophysical exploration, Pt. 1. — The magnetic 
airborne detector AN/ASQ-3A was developed 
by the Naval Ordnance Laboratory and the 
Bell Telephone Laboratories and was produced 
by the Western Electric Co. for use in the U. S. 
Navy's Bureau of Aeronautics for locating 
submarines from aircraft. Modified systems of 
this type have been used extensively by the 
Naval Ordnance Laboratory and the U. S. 
Geological Survey in airborne geophysical sur- 
veys throughout the United States and in 

In the magnetometer described, detection of 
a magnetic field employs an inductor with 
nearly ideal magnetization and saturation 
characteristics and negligible hysteresis. When 
the inductor winding is driven well beyond 
saturation field for the core material by a sine- 
wave current, the resulting distortion of flux 
wave-form produces harmonics of the driving 
frequency. The even-harmonic content is pro- 
portional to the magnetic field applied along 
the inductor axis. The second harmonic is 
selected bj r suitable filtering. Three such induc- 
tors are arranged mutually perpendicular. Two 
of these are used in a servo system to maintain 
automatically the axis of the third along the 
direction of the total field, and also to com- 
pensate for flight errors in the orientation 
system. A controlled variable direct current 
is supplied to the third inductor winding to 
suppress the total field to a small fraction of the 
earth's field. The low-frequency response of the 
system permits continuous recording of the 
earth's magnetic field intensity and anomalies 
as small as 10~ 5 gauss (1 gamma) while air- 
borne. (Author's abstract.) 

F. Keller, U. S. Geological Survey: Air- 
borne magnetometer in geophysical explorations, 
Pt. 2. — The adaptation by the Geological 
Survey in cooperation with the Naval Ordnance 
Laboratory of the AN/ASQ-3A Magnetic 
Airborne Detector to problems of exploration 
in geophysics has made it possible to produce 

Jan. 15, 1948 

proceedings: philosophical society 


maps of large areas showing total intensity 
variations of the earth's magnetic field with 
greater accuracy and speed and with less ex- 
pense than by ground measurements. 

The U. S. Geological Survey obtains the 
data by towing the magnetic detector element 
behind a twin-engined airplane in a stream- 
lined housing called a "bird." Variations in 
magnetic intensity are recorded on a tape in the 
plane. The location of the plane is determined 
at all times by a gyroscopically stabilized 
Sonne continuous strip camera. A system of 
electrically operated counters and edge mark- 
ers make it possible to correlate recognizable 
ground check points with the strip film and 
magnetometer record. 

Magnetic surveys of over 120,000 square 
miles in areas which covered extreme contrasts 
of geological and magnetic environments have 
proved the value of the airborne magnetometer 
as a reconnaissance tool in exploration for 
mineral resources. (Author's abstract.) 

1277th Meeting 

The 1277th meeting was the occasion of the 
Sixteenth Joseph Henry Lecture and was held 
in the auditorium of the U. S. National Mu- 
seum, February 15, 1947, President Mohler 

Program: Jesse W. Beams, Professor of 
Physics in the University of Virginia: High 
centrifugal fields. This address has been pub- 
lished in full in the Journal 37: 221-241. July 

1278th Meeting 

The 1278th meeting was held in the audi- 
torium of the Cosmos Club, March 1, 1947, 
President Mohler presiding. 

Program: U. Fano, National Bureau of 
Standards: Contributions of physics to biology, 
Pt. 1. — The borderland between physics and 
biology is today the subject of widespread 
interest, especially from the side of young 
physicists, who think of making it their main 
line of work. It seems timely, therefore, to 
examine critically which specific, actual or 
potential, lines of endeavor should go under the 
name of biophysics. A few lines are well 
defined, such as: (a) the design, production, 
and employment of physical tools which help 
in solving biological problems; this is primarily 
a branch of applied physics; (b) the investiga- 

tion of biological problems which involve phys- 
ical measurements and tools; this requires 
knowledge of physical apparatus and tech- 
niques belonging primarily to biology as far as 
its motivation and methodology are concerned; 
(c) the participation of workers with training 
and experience in physics in the investigation 
of biological problems; this has often proved 
extremely effective. 

It should be stressed that the field of en- 
deavor is a biological one and physics plays no 
further part than lending ways of thought and 
methods of approach to problems. (Author's 

G. Gamow, The George Washintgon Uni- 
versity: Contributions of physics to biology, Pt. 
2. — Until comparatively recent time the phe- 
nomena of life were so widely separated from 
familiar physical and chemical phenomena in 
inorganic matter that it seemed that the gap 
between the two could never be covered. This 
led to various "vitalistic" theories which in- 
sisted that the phenomenon of life is principally 
different from the ordinary physical phenom- 
ena and requires for their explanation some 
peculiar force known as "vis vitalis." 

The recent progress of biology seems to 
cover the existing gap. In fact, the things like 
viruses and genes seem to represent the "missing 
links" between organic and inorganic matter. 
Although there is little doubt that viruses and 
genes represent the ordinary (although very 
complex) chemical molecules obeying the or- 
dinary laws of quantum chemistry, they are 
also attributed to all basic properties of living 

It is time now to attack the problem of these 
elementary living organisms from the purely 
physical point of view, and, while the biolo- 
gists dig deeper and deeper into the properties 
of these entities from the complex and of bio- 
logical forms, the physicist may be expected to 
approach the problem from the other end and 
"build up" the theory of gene and virus mole- 
cules on the basis of our present vast knowledge 
concerning the properties of atoms and atomic 
combinations. Just as the hydrogen atom 
represents the simplest model of atomic struc- 
ture, and the deuteron the simplest nucleus, 
virus-particles are most probably the simplest 
biological models which can be treated theo- 

Although there are still tremendous difficul- 



ties ahead, we may not be so far from the pos- 
sibility of constructing the model of the sim- 
plest living unit on the basis of purely atomic 
considerations. (Author's abstract.) 

1279th Meeting 

The 1279th meeting was held in the audi- 
torium of the Cosmos Club, March 15, 1947, 
President Mohler presiding. 

Program: S. E. Forbush, Department of 
Terrestrial Magnetism: Cosmic rays from the 
sunt — During the past 10 years three unusual 
and sudden increases in cosmic-ray intensity, 
lasting several hours, have been observed 
simultaneously in different parts of the world 
except at the Equator where no increase was 
observed. All three increases began nearly 
simultaneously with unusually long and intense 
radio-fadeouts and solar-flares. The cosmic-ray 
increases were similar and simultaneous on the 
day and night side of the earth. Magnetograms 
from several magnetic observatories indicate 
that the increase in cosmic-ray intensity can 
not be ascribed to changes in the earth's mag- 
netic field. While the increases in cosmic-ray 
intensity might be ascribed to changes in the 
sun's general magnetic field which would per- 
mit more cosmic rays from outer space to reach 
the earth, such mechanism should be equally 
effective whether the solar-flare occurred in the 
front or back side of the sun. This with other 
arguments indicates that changing magnetic 
fields associated with a sunspot or flare may 
act as a magnetic accelerator for charged par- 
ticles. Similar processes on stars might account 
for all cosmic rays. (Author's abstract.) 

P. H. Abelson, Department of Terrestrial 
Magnetism: The transuranic elements. — The 
paper reviewed the facts leading to the dis- 
covery of neptunium. The nuclear and chemi- 
cal properties of the transuranic elements were 
discussed. Possible methods of producing ele- 
ments of higher atomic number than 96 were 
described. These included the bombardment 
of uranium or plutonium with high speed 
carbon, oxygen, or other nuclei. (Author's ab- 

1280th Meeting 

The 1280th meeting was held in the audi- 
torium of the Cosmos Club, March 29, 1947, 
President Mohler presiding. 

Program: Hugh L. Dryden, National Bu- 
reau of Standards : Exploring the fundamentals 

of aerodynamics. — This paper has been pub- 
lished in full in the Journal37: 145-156. May 

1281st Meeting 

The 1281st meeting was held in the audi- 
torium of the Cosmos Club, April 12, 1947, 
President Mohler presiding. 

Program: A. H. Bennett, American Optical 
Co. : The phase contrast microscope. — The older 
microscopic methods are particularly useful for 
certain types of specimens, but for materials 
whose details have little absorption differences 
but differences in optical path, the phase 
method has advantages. In phase microscopy 
such microscopic details are imaged as bright- 
ness differences to which the eye or photo- 
graphic plate are sensitive. Work leading to 
phase microscopy was done by Abbe* (prior 
to 1892), Bratuschek (1892), Rheinberg (1904), 
and Conrady (1905). Zernicke (1935) de- 
scribed the application of the phase contrast 
method to microscopy. Burch and Stock 
(1942) described their results using simplified 
equipment. Study of the theory, application, 
and apparatus for the method was begun by 
Richards, Osterberg, Jupnik, and the writer 
in 1941. In phase microscopy an annular ring 
placed near the lower focal plane of the sub- 
stage condenser acts as a secondary light 
source. This ring is imaged by the condenser 
and the objective onto a diffraction plate which 
by means of evaporated metal and dielectric 
films, alters the amplitude and phase relation- 
ships between the light undeviated by the 
specimen and the diffracted light. The light 
distribution in the image formed under the 
above conditions is determined by means of 
diffraction theory. (Author's abstract.) 

O. W. Richards, American Optical Co.: 
Applications of the phase contrast microscope. — 
While theoretical considerations indicate the 
possibility of a large number of diffraction 
plates, a relatively small number serve the 
needs of practical microscopy. Bright contrast 
is useful in counting and for finer details. 
Dark contrast is useful and sometimes is pre- 
ferred as the image resembles that from stained 
preparations. Detail is brought out in living 
cells, emulsions, and on transparent surfaces. 
Photomicrographic records are possible and 
stereophotomicrographs reveal detail in three 
dimensions. When stained or pigmented, speci- 

Jan. 15, 1948 

proceedings: philosophical society 


mens do not have high contrast, it is possible 
to add phase contrast of greater visibility. All 
four types of Spencer diffraction plates have 
proved of use in microscopy. Replicas of Form- 
var, acetate, or silica surfaces may be examined 
with the phase microscope which opens a new 
field for investigation. Photomicrograph lan- 
tern slides were shown of typical applications 
including living unstained epithelial cells, 
bacteria, fungi, protozoa, and spermatozoa; 
unstained tissues, chromosomes, and malaria 
parasites; stained chromosomes, urine casts, 
emulsions, wool, rayon, paper, and replicas of 
steel surfaces. (Author's abstract.) 

W. R. Duryee, Department of Terrestrial 
Magnetism: A film on the mechanism of cell 
division. — The paper consisted essentially of 
the showing of a film prepared by K. Michel 
of the Zeiss Co. using phase contrast lenses and 
time-lapse photography. 

The material photographed consisted of 
living spermatocytes of the grasshopper Pso- 
phus stridulus Linnaeus isolated in a physiolog- 
ical salt solution. Since time lapse photog- 
raphy was used, processes lasting over 5 to 
15 hours were reviewed in a few minutes. In 
addition the phase contrast lenses brought out 
many details, relationships, and structural 
contrast that the human eye unaided could not 
see in normally transparent cells. 

Major sequences in the film were, first, 
prophase changes in the nucleus. Chromosomes 
were visible shortening and thickening in their 
colloidal matrix. Arrangement of chromosomes 
on the spindles was followed by grouping of 
mitochondria in the mid-cytoplasm, next in the 
metaphase details of spindle lateral spread 
preceded actual elongation. Thirdly, separation 
of the chromosomes and other cytoplasmic 
movements during anaphase were shown. Final 
stages of cytoplasmic fission and telophase 
reconstitution of the daughter nuclei were 
clearly photographed. 

A remarkable sequence showed continual 
elongation of spindle substance even after each 
cell had divided. Concomitant sol-gel reversals 
of the polar surfaces were evident. New details 
of spermatocyte transformation into the early 
spermatic concluded the picture. 

The film has been adopted by the Surgeon 
General's office as a medical technicians' train- 
ing film and was shown through the courtesy 
of the War Department. (Author's abstract.) 

1282d Meeting 

This was a joint meeting with the Washing- 
ton Academy of Sciences, held in the audito- 
rium of the Cosmos Club, April 17, 1947. 

Program: P. W. Bridgman, Professor of 
Physics in Harvard University. — High pres- 
sures and their effects. 

1283d Meeting 

The 1283d meeting was held in the audi- 
torium of the Cosmos Club, April 26, 1947, 
President Mohler presiding. 

Program: Paul F. Wacker, National Bureau 
of Standards: Heat capacities of gases. — Ex- 
perimental heat capacity data for gases are 
useful both for the calculations of heat require- 
ments in industrial operations and in checking 
statistical calculations of heat capacity, heat 
content, entropy, and free energy of compli- 
cated molecules. These calculations are made 
for extremely wide temperature ranges and 
permit calculations of yields o'f reactions as 
well as heat requirements. 

Heat leaks were made negligible by the use 
of thermostated radiation shields and a care- 
fully designed calorimeter. This absence of 
heat leaks made high precision possible and re* 
duced consistent errors to very small values. 
The working time was also reduced, since or- 
dinarily it is necessary to make measurements 
as a function of flow rate in order to correct for 
heat leak. As a check for consistent errors, the 
heat capacity of oxygen was measured and 
compared with highly accurate values calcu- 
lated statistically from spectroscopic data. The 
values at -30°C and +90°C differed from the 
spectroscopic data by 0.06 percent and 0.10 
percent, respectively, while the experimental 
value at -|-40 o C did not differ from the spec- 
troscopic data by a significant amount. Con- 
sideration of the data obtained for oxygen and 
the known sources of error indicates that the 
measurements with the apparatus have a 
probable error of 0.06 percent or 0.07 percent. 
(Author's abstract.) 

W. H. Avery, Johns Hopkins Applied 
Physics Laboratory: Infrared spectra at low 
temperatures. — Difficulties caused by overlap- 
ping of the infrared absorption bands restrict 
the application of these spectra to the analysis 
of complex hydrocarbon mixtures. Theoretical 
considerations suggested that these difficulties 



VOL. 38, XO. 1 

might be minimized by observing the spectra 
at low temperatures, since the width of the 
absorption bands is predicted to be propor- 
tional to the square root of the absolute tem- 
perature. Using a cell with a reflecting bottom 
surface which could be immersed in liquid 
hydrogen, upon which films were condensed, 
spectra were obtained of a number of saturated 
hydrocarbons. The spectra were found to show 
the predicted narrowing of the absorption 
bands and a striking decrease in overlapping of 
the bands. Quantitative application of the 
method was interfered with, however, because 
of the scattering due to small crystals of the 
condensed hydrocarbons. A change from the 
gaseous to solid phase did not produce any 
significant shifts in the position of the bands of 
saturated hydrocarbons, but there were some 
indications that overtone and combination 
bands were changed in intensity by the conden- 
sation. The use of liquid helium was suggested 
for a further improvement in resolution of the 
spectra. A theoretical analysis of probability 
of the overlapping bands in complex mixtures 
showed that the use of low temperatures might 
be expected to increase tremendously the num- 
ber of mixtures susceptible to infrared analysis. 
{Author's abstract.) 

1284th Meeting 

The 1284th meeting was held in the audi- 
torium of the Cosmos Club, May 10, 1947, 
President Mohler presiding. 

Program: D. H. Menzel, Harvard College 
Observatory : Problems of the sun. — Our rapidly 
accumulating observational data on the sun 
have served to focus our attention on the many 
unsolved problems of the solar atmosphere. 
Although the surface layers appear to possess 
a temperature of about 6,000°C, the tempera- 
ture and excitation rise rapidly in the higher 
levels. In the upper chromosphere and promi- 
nences, the temperatures range from 10,000°C 
to 25,000°C. In the solar corona, temperatures 
may attain values in excess of a million de- 
grees. There are many additional problems such 
as the support of the prominences and chromo- 
spheric layers and the explanation of the gases 
in the solar atmosphere. Thus far, we have no 
completely satisfactory theory of any of these 

The solar observations made regularly at the 
High Altitude Observatory of the Harvard 
Universit} r and the University of Colorado, 
located at Climax, Colo., have provided the 
astrophysicists with new indices of solar ac- 
tivity. Even though we do not yet understand 
the basic physical processes, we realize that 
marked changes in the output of solar ultra- 
violet radiation accompany the prominences 
and coronal disturbances. These ultraviolet 
variations, in turn, affect the ionosphere, the 
earth's magnetic field, and the luminosity of 
the night sky. There is also a high probability 
that there may be a relationship between solar 
activity and terrestrial weather. The ozone 
layer of the earth's atmosphere acts as a me- 
dium to transmit solar impulses from the upper 
to the lower atmospheric levels. 

Study of the sun, for the above-mentioned 
practical reasons, assumes an increasingly im- 
portant place in science. Accurate forecasting 
of radio-propagation conditions or of weather in 
general would possess immense economic value. 
For example, if only we could tell the farmer 
that the next growing season will be hot or 
cold, wet or dry, late or early, many losses 
would be avoided. 

For the long-range view, study of solar radi- 
ation has additional practical applications. No 
natural process is now replacing the stock of 
oil and coal, produced over geological ages. 
An understanding of the sun and solar process- 
es will help us to utilize more effectively the 
solar radiation, upon which we shall eventually 
have to depend for heat and power. (Author's 

1285th Meeting 

The 1285th meeting was held in the audi- 
torium of the Cosmos Club, May 24, 1947, 
President Mohler presiding. 

Program: A. E. Ruark, Johns Hopkins 
Applied Physics Laboratory : Behavior of par- 
ticles in scattering and absorbing media. 

A. J. Shneiderov presented an informal 
communication regarding gravitation. 

The president announced the election of the 
following Committee on Communications for 
the season 1947-48: G. B. Sabine, chairman, 
R. C. Herman, and Thomas Carroll. 

Walter L. Cheney, Recording Secretary. 



Mathematics. — The role of the concept of equivalence in the study of 
physical and mathematical systems. Richard Stevens Buring- 

TON 1 

Botany. — Studies in Lonchocarpus and related genera, II: Miscellaneous 
Middle American Lonchocarpi. Frederick J. Hermann 11 

Zoology. — On the crayfishes of the Limosus section of the genus 
Orconedes (Decapoda: Astacidae). Horton H. Hobbs, Jr 14 

Ornithology. — The races of the black-throated sunbird, Aethopyga 
saturate (Hodgson). H. G. Deignan 21 

Proceedings: The Academy 23 

Proceedings: Philosophical Society 25 

This Journal is Indexed in the International Index to Periodicals 

Vol 38 

February 15, 1948 

No. 2 





James I. Hoffman 



Alan Stone 


Frank C. Kracek 



Lawrence A. Wood 


J. P. E. Morrison 


Elbert L. Little, Jr. 


Richard E. Blackwelder 


James S. Williams 


Waldo R. Wedel 


Irl C. Schoonover 




450 Ahnaip St. 

at Menasha, Wisconsin 

Entered as aeoond class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a special rate of postage provided for in the Aot of February 28, 1025. 

Authorised January 21, 1933, 

Journal of the Washington Academy of Sciencea 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author' 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate) : 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20- 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 .50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8.00 0.60 

Vol. 23 to current vol. (12 numbers per vol.) 7.50 0.90 

• Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers. 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 




Vol. 38 

February 15, 1948 

No. 2 

PHYSICS. — The measurement of high voltage. 1 
Bureau of Standards. 

F. M. Defandorf, National 


This paper presents an outline of the 
basic principles used in measuring high 
voltages. It appraises the accuracies at- 
tained by some of the methods and devices 
currently in use, rather than attempts to 
give historical credit for discovery and in- 
vention, although numerous references are 
cited which should prove useful to one inter- 
ested in tracing such matters. The expres- 
sion "high voltage," in the title of this paper, 
will be construed to include voltages of the 
order of 100 kilo volts and higher. It will be 
convenient to leave out of consideration 
many voltage measuring devices useful in 
the neighborhood of 100 kv and lower but 
of doubtful value in extending the range to 
much higher voltage by increasing the pro- 
portions of the device. This demarcation 
also serves to exclude from this discussion 
the methods of measurement of voltage at 
extremely high frequencies because voltages 
of this magnitude are nonexistent at those 
frequencies. The first portion of this paper 
will logically be devoted to a discussion of 
certain reliable methods of extension in 
range from low voltage standards. 

It is proper to ask why one should be in- 
terested in the accurate measurement of 
high voltage. One must be interested be- 
cause many phenomena depend on voltage. 
For instance, the hardness limit of X-radia- 
tion emitted by an X-ray tube depends 
upon the maximum value of the voltage 
across the tube including the ripple or any 
superposed surge. Similarly the sparkover 
voltage of a sphere gap and the a-c flash- 

1 Address of the Retiring President of the 
Philosophical Society of Washington, delivered at 
the 1274th meeting of the Society on January 
4, 1947. Received July 28, 1947. 

over values for a string of line insulators 
depend upon the crest of the alternating 
voltage applied to the gap or string. In 
those cases it is the maximum voltage aris- 
ing from the combination of all voltage 
components in which we are interested. In 
the surge-voltage breakdown test of a light- 
ning arrester it is the crest of the applied 
surge that is fundamental, although the 
form of the surge, if it rises very rapidly, 
may be important. This is because with 
surges of increasingly steep wave front the 
crest breakdown voltage actually increases 
because it requires an appreciable though 
short time to establish the mechanism of 
breakdown. On the other hand, in the com- 
mercial sale of energy it is the effective value 
of the voltage and its phase with respect to 
the current that is important. Thus the 
measurement of high voltage in the cases 
just mentioned presupposes some knowl- 
edge not only of the magnitude of voltage, 
say from thousands to several millions of 
volts, but also a knowledge of the variation 
of the voltage with time. This variation of 
voltage with time, which may have the 
form of a wave, pulse, ripple, or some other 
shape in no way resembling our everyday 
conception of a wave, is called wave form. 
From the foregoing discussion one may 
conclude that the significant characteristics 
of voltage to be measured, illustrated in 
Fig. 1, may be listed as: 

1. Direct voltage (d-c voltage). 

2. Effective alternating voltage (effective a-c 

3. Average alternating voltage (average a-c 

4. Crest or maximum voltage of a rippled 
d-c-, alternating-, or surge-voltage. 

5. The wave form of a surge- or alternating- 




For the measurement of voltage it has 
been found convenient to adopt as a pri- 
mary standard, the voltage of the standard 
cell. The unsaturated standard cell, the 
voltage of which is approximately 1.0186 + 
volts, is almost universally used as a lab- 
oratory reference standard for voltage 
measurements. When kept in a thermo- 
stated enclosure and used under favorable 
conditions, it exhibits a remarkable con- 
stancy of voltage. The small differences that 
may arise in a thermostated group of cells 
from year to year are of the order of micro- 
volts. Unfortunately, similar electrochem- 
ical devices having 100, 1,000, or 1,000,000 
times the voltage of a standard cell do not 
exist. Therefore, in the usual measurement 
of higher voltages, it has been necessary to 
develop devices which ultimately refer back 
to the standard cell. 

Resistance methods of extending the 
range of voltage upward from the value of 
the voltage of the standard cell make use of 
the fact that by selecting suitable alloys and 
heat treating them properly resistors can be 
constructed to have a resistance, or a ratio 
of resistance, that remains constant to 
within a few parts in 100,000 over a satis- 



factory temperature range. The use of a 
potentiometer or of a tap resistor (volt box) 
is the simplest method for extending the 
range of measurements of direct voltages by 
reference to the standard cell. If r is the 
resistance from one end to the tap point and 
R is the total resistance including r, then 
the voltage e across r can be measured with 
reference to a standard cell by using a 
potentiometer and the voltage V across R 
is then V = e- R/r. 

Although this simple resistance method 
is adequate for small steady direct voltages 
it requires modification for the precise 
measurement of the high varying direct-, 
alternating-, surge-, and pulse-voltages that 
one encounters in practice. The manner in 
which the voltage at any instant varies 
with the time is a fundamental factor in the 
problem of measurement. It is necessary to 
express alternating voltage in terms of di- 
rect voltage which can be evaluated by 
reference to the standard cell. Instruments 
designed to read the same on the effective 
value of alternating voltage as on direct 
voltage are called ''transfer instruments. " 
As transfer instruments, electrostatic in- 
struments (1) are favored in Great Britain, 

V nvP=.637 V 



Fig. 1. — Wave forms. 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


while electrodynamic instruments (2) find 
favor in this country. An electrostatic volt- 
meter, for instance, should give the same 
indication for a direct voltage V as it does 
in the case of an alternating voltage of pure 
sine-wave form for which the crest or maxi- 
mum voltage V m = y/2 V. 

If rectified alternating voltage is applied 
to a direct current voltmeter the instru- 
ment will read the average value of voltage 
applied to it, i.e., V a ve=Q-/t)Jl vdt. 

If, however, completely rectified alter- 
nating voltage is applied to a good electro- 
static voltmeter the voltmeter will read the 
crest voltage. Crest voltage can also be de- 
termined from the length of a spark gap 
across which it will just cause a discharge, 
or it can be readily evaluated from the wave 
form determined by use of a calibrated os- 

A more detailed knowledge of the varia- 
tion of the voltage with time such as is 
given by an oscillograph becomes of special 
interest in the case of pulse and surge volt- 
ages. Thus it would seem appropriate to ap- 
praise the value of the several devices used 
in the delineation of wave form, and to men- 
tion each type under the particular voltage 
divider or device with which it is generally 

A consideration of high- volt age measur- 
ing devices appears to lend itself better for 
discussion under a classification of methods 
of measurement or types of devices rather 
than under the classification of character- 
istics of voltage previously outlined. The 
devices which are useful in measuring high 
voltage may be conveniently classified as 
to type by considering whether the method 
of measurement employs : 

1. A high series impedance with a low-imped- 
ance instrument to indicate current through the 

2. A potential divider in which a fraction of the 
total voltage is measured across tap points of 
the impedance. 

3. A voltage transformer that permits measure- 
ment of a low voltage having a direct ratio to the 
high voltage. 

4. A generating voltmeter in which a voltage 
proportional to the field intensity in the region of 
the instrument is indicated. 

5. A spark discharge in which the length of the 
spark gives a measure of the voltage. 

6. The cooling effect of an "electric wind" as in 
the ionic wind voltmeter. 

7. Force arising from the attraction or repul- 
sion of electrostatic charges on electrodes. 

8. The deflection of a stream of charged par- 
ticles by means of a known field, either electro- 
static or magnetic, after their acceleration in 
vacuo by the voltage to be measured. 

The devices used in measuring high volt- 
age will be considered in accordance with 
the above classification and in the order 


The simplest method of measuring high 
voltage would appear to be to connect a 
high impedance in series with a sufficiently 
sensitive current measuring instrument hav- 
ing a negligible impedance compared to the 
value of the high series impedance, Z. Val- 
ues of the indicated current, », would then 
give the high voltage, V = iZ. The imped- 
ance Z may, of course, be primarily resis- 
tive, inductive, or capacitive or combina- 
tions of these elements. 

Series resistors. — This basically simple 
method has been widely used in nuclear dis- 
integration work for direct voltage measure- 
ments, wherein the impedance Z is built of 
many high resistance units in series (dia- 
gram at A in Fig. 2), care being taken to 
insure that the current entering at the high 
voltage end of the resistor is the same as 
that leaving through the deflecting instru- 
ment at the low voltage end. For reliable 
measurements it is necessary to be sure that 
the electrical leakage across insulating sup- 
ports of the resistor and from section to sec- 
tion is negligible and that there is negligible 
corona current from the units. Changes 
arising from self-heating must be made 
negligibly small or must be allowed for. The 
design of such a series resistor should be 
substantially the same as for the potential 
divider type to be discussed later in detail. 

Series reactors. — For alternating voltage 
measurements, series reactors have been 
used as the series impedance (3) . They have 
the drawback, however, of requiring iron 
cores at lower frequencies if the inductive 
reactance is to be made large in comparison 
with the resistance of the windings. Stray 
and distributed capacitance effects raise 
additional objections so that series reactors 
have been little used in high voltage meas- 








/j Am 

V = 


m Am 

A B 

Fig. 2. — Diagrams of series impedors. 

Series capacitors. — For the measurement 
of high alternating voltage, capacitance 
may be used in series with the low voltage 
indicating instrument, as indicated at B 
and C in Fig. 2. The capacitor, which must 
be relatively free from losses, corona, and 
brush discharges, magnifies the effect of 
harmonics present in the voltage wave form 
unless the low voltage instrument is of the 
capacitance type. 2 However, if harmonics 
are present, corrections can be applied. 
This device in a form for measuring high- 
crest voltage was originally described by 
Chubb (4) who employed the two spheres 
of a sphere gap as the capacitor, the lower 
and grounded sphere being insulated to 
permit rectification of the capacitance cur- 
rent to it. Such devices (5, 6, 7) appear to 
have been used more generally in foreign 
laboratories than in this country. The ar- 
rangement is indicated at C in Fig. 2. Hae- 
fely & Co. patented (7) such a device which 
employs as one electrode of the high-volt- 
age capacitor, a large insulated circular 
segment of the lower sphere. The remainder 
of the sphere serves as the grounded guard 
for this segmental electrode. The upper 
sphere serves as the high-voltage electrode. 

2 Here the combination serves as a potential 
divider and a low voltage electrostatic voltmeter 
across a large capacitance gives effective values 
of voltage independent of wave form. 

These devices generally permit only a 
rough computation of capacitance and 
therefore need to be calibrated against some 
other voltage standard, but they have the 
advantage over the sphere spark gap of giv- 
ing a continuous rather than transient indi- 
cation of voltage when used with an elec- 
trostatic voltmeter or rectifier-milliam- 
meter combination. Although the early de- 
vice of Chubb gave the crest value 3 of volt- 
age, later similar arrangements (9) were 
devised that permit the determination of 
both crest- and effective-values of voltage, 
and when supplemented with a cathode ray 
oscillograph or synchronous commutator 
(8) give the high voltage wave form as well. 
For relative measurements, this series ca- 
pacitance method should be good to a few 
tenths of one percent. 


The potential divider is essentially some 
form of impedance with one or more tap 
points permitting the measurement of the 
voltage drop between tap points by a 
method which preferably does not change 
appreciably the current flowing through the 
divider. The potential divider is connected 
across the voltage to be measured. The 

3 Except in the case of alternating voltages of 
unusual wave form in which there are several 
maxima (8). 

W' ■ '■■■■ : W 


1 B 

8 ? R*" af 

B : '-^w- 'Jr - "^P* *"*^ - JT^ 

K' "■ IS 

F. M. Defandorf 
President of the Philosophical Society of Washington, 1946 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


value of voltage measured between taps, 
when multiplied by the ratio of the value 
of the total impedance to that of the tapped 
section then gives the value of voltage 
across the total impedance. Potential di- 
viders are commonly constructed of sections 
of resistance, of resistance shunted by 
capacitance, or of capacitance. Types of 
each of these will be described. 

Corona shielded resistor. — For a resistive 
divider one may use n resistors, each of re- 
sistance r, all exactly alike (or nearly so), 
connected in series, or one resistor (n— 1) 
times as large as the smaller one connected 
in series with it, and across which the volt- 
age drop is measured. For use at higher 
voltages, both the electrical and mechanical 
designs of the physical device require care. 
Improvements in his earlier designs (10) 
have been incorporated by L. S. Taylor in a 
resistor, not yet described, which is in 
use in the measurement of the high rec- 
tified voltage supply for a 1.4-million- 
volt X-ray tube in the X-ray Labora- 
tory of the National Bureau of Standards. 
As seen in Fig. 3, the column at the left 
side contains the corona-shielded resistor, 
which is installed immediately adjacent 
to the column of 10 large cascaded-rectifier 
tanks, each of which at maximum volt- 
age contributes 140 kv to the total voltage 
of 1.4 million. This arrangement provides 
double shielding. The external system 
seen in Fig. 3 has tubular connections 
from the spun-metal corona shields to 
each section of the cascaded supply. The 
internal resistor assembly, also separately 
corona shielded, is thoroughly insulated 
from this outer system. Thus the currents 
required by the outer shields are supplied 
as a sectionalized direct load on the supply 
independently of the measuring circuit. 
This outer shield therefore takes care of 
current flow arising from external ioniza- 
tion, migratory dust particles, and minor 
surface discharges as well as surface leakage 
across the insulating columnar supports, 
and relieves the internal resistor of the 
resulting irregularly varying effects. In 
addition, the separate outer shields provide 
mutual field grading and the large top 
shield is effective in shaping the general 
electrostatic field. The internal resistor as- 

sembly is sectionalized, each section con- 
sisting of uniformly spaced wire- wound re- 
sistors located between toroidally shaped 
corona shields, which are electrically con- 
nected to the end of each group of resistors. 
Isolantite insulation is used for supporting 
the resistors and the uniformly spaced co- 
rona shields. The shields were constructed 
of smooth copper-tubing of circular section 
in order to provide satisfactory field grading 
along the resistor and thus to insure free- 
dom from corona discharge currents at the 
highest intended operating voltage. The de- 
sign provides for adequate insulation, me- 
chanical support, and centering of this 

Fig. 3. — 1.4-million-volt rectifier column with 
shielded resistor column on left side. 



VOL. 38, NO. 2 

column inside the outer corona shield. Each 
section was adjusted and thoroughly tested 
at rated voltage before assembly and subse- 
quent measurements of individual sections 
show their resistance to be practically free 
from drift. These precautions were taken to 
make sure that the current throughout the 
resistor will be the same as at its low voltage 
end where measurement of the current is 
made; i.e., in order that the high voltage, 
V, being measured is equal to nri where nr 
is the total resistance of the stack of n 
shielded resistors. Actually the current, i, 
is often determined by adding a small re- 
sistor, r , and measuring v = ir„ by means of 
a potentiometer. 

Manganin wire-wound resistors and a 
relatively large current of 1 milliampere 
were chosen as design values in order to 
minimize parasitic effects of leakage, ab- 
sorption, and corona currents, and thus to 
insure a resistor that would permit measure- 
ments to 0.1 percent or better. A current of 
1 milliampere corresponds to a load of 1.4 
kilowatts, so that the resistor was designed 
to dissipate a corresponding amount of 
heat. Carbon or the so-called "metallized" 
radio-type resistors have a much higher 
temperature coefficient and are in general 
less stable than wire-wound resistors. Con- 
siderable care must therefore be taken in 
their use to insure freedom from thermal 
and voltage effects. Although reduction in 
self-heating and in energy loss in such resis- 
tors may be secured by reducing the current 
say to 0.1 milliampere, the effects of leakage 
currents become relatively more important 
and may contribute to some extent in re- 
ducing the attainable accuracy to the order 
of 0.5 percent. 

Still another consideration that should 
not be overlooked is the shunting effect on 
the resistance of the capacitance of the 
shields. In the case of a resistor of the high 
resistance type, sudden surges are easily 
transmitted through the shunting capaci- 
tances so that the measuring instrument, 
which must be correspondingly more sen- 
sitive, requires increased surge-protection. 
Thus there are drawbacks to either in- 
creasing the current in the measuring resis- 
tor to values greatly exceeding one milli- 
ampere because of increased energy loss, or 

decreasing it to values much below 0.05 
milliampere because of the instability of the 
resistance insulation, and of other troubles. 
For much of the pioneering nuclear disinte- 
gration work (11) carbon resistors available 
in units of much higher resistance per unit 
length than the wire-wound type have 
proved both economical and useful. 

It should be observed that the relatively 
high shunting capacitance of the corona 
shields of this type of resistor limits its 
usefulness as a resistor for alternating cur- 
rent measurements. 

Capacitance Melded and guarded resistors 
for alternating current. — The resistor di- 
vider, in high favor for the measurement 
of direct voltage, has certain specific defects 
in addition to its relatively large consump- 
tion of power. It necessarily has residual 
inductance and stray capacitance associ- 
ated with its resistance. As shown at A in 
Fig. 4, each section n • • • r n may be con- 
sidered as having at its terminals a lumped 
shunting capacitance of value & • • • C n , 
and between its upper terminal and ground 
a capacitance C a • • • C m . If the value of 
capacitive reactance l/27r/Ci is ^>n and 
l/2TrfCa2>nri and if the impedance used in 
the measuring circuit shunting r n is ^>r n , 
the voltage division by resistance will be 
trustworthy. As a result of the increase in 
shunting effect of the stray capacitances 
with frequency, these inequalities grow 
less until the division of the applied voltage 
is no longer proportional to resistance. 

The shielded a-c resistor (12) minimizes 
the effects of the unavoidable ground ca- 
pacitances at the cost of additional energy 
dissipation, by employing a second or guard 
resistor in parallel with the first or " work- 
ing" resistor. This guard resistor is so con- 
nected to the shields that it supplies current 
to the ground capacitances that otherwise 
would have to be charged through and 
along portions of the working resistor. In 
an a-c shielded resistor the working, or 
shielded resistor, is composed of sections 
of value r each contained within a metal 
shield or box as indicated at B in Fig. 4. 
Each shield is maintained at a potential 
corresponding to the midpoint of its en- 
closed resistor hv connecting it to a tap 
point on the proper section R n of the guard 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


resistor. Thus this arrangement makes the 
value of each shunting capacitance definite 
and supplies the ground capacitance cur- 
rent from the guard circuit. This is ac- 
complished at the cost of a slight increase 
in the shunting capacitance of the indi- 
vidual resistors because of the capacitance 
to the shield in which they are located, 
since one end of the enclosed resistor is 
above and the other below the potential of 
the shield. However, because the potential 
difference across the capacitance of either 
end of the resistor element to its shield is 
only one-half the voltage drop in this sec- 
tion of resistance r, the shunting effect 
need not be excessive at low frequencies, 
say 25 or 60 cycles. Such a resistor, com- 
posed of about 25 shielded sections, is used 
in a voltage transformer testing setup for 
phase-angle measurements and ratio meas- 
urements to 0.01 percent at the National 
Bureau of Standards. This resistor is rated 
at 0.05 amperes in each of the two (working 
and guard) 500,000-ohm circuits. In this 
device, which at 25 kv absorbs 2.5 kw, the 
individual sections of the working resistor 
are enclosed in metal shield boxes, which 

are filled with oil in order to improve ther- 
mal characteristics. This resistor is ar- 
ranged in tiers, each tier consisting of four 
working resistor boxes with corresponding 
sections of the guard resistor supported on 
a mahogany framework. Tiers are as- 
sembled one above the other by means of 
porcelain bus-type insulators, which serve 
to provide additional insulation and to 
permit bolting together of the resistor- 
supporting frameworks to form a mechan- 
ically stable assembly. Each box contains 
20 flat mica cards wound unifilarly with 
manganin wire. It is not feasible to extend 
the range of a-c shielded resistors and retain 
the same order of accuracy (0.01 percent) 
by this procedure much above 30 kv because 
of the effects of the capacitances which 
shunt the guard resistances. The resulting 
error increases as the fourth power of the 

An a-c shielded resistor of this type is 
also satisfactory for use as a series resistor 
in conjunction with an indicating volt- 
meter or sensitive oscillograph. It may be 
used equally well on direct current but 
would ordinarily not be used because its 


r '\ 

K C a 
! 7°. 


\ C b 

i 7 C 2 


;/ C 3 




r n t C n 


r = r ,=r 2 = r 3 --r n 

Fig. 4. — A-C shielded resistor. 







R 2 =r 

R 3 =r 

R n=2 r 




to CRO 

Fig. 5. — Surge resistor. 

load on the source is greater than that for 
the simpler corona shielded resistor. 

Weller (13) devised a shielded a-c resistor 
for use in transformer testing up to 132 kv 
in which the shield current was supplied by 
auto-transformers rated at 75 kva, instead 
of by a tapped resistor. 

Surge resistors. — In surge voltage testing 
equipment used for studies of surge effects 
on electric power and transmission equip- 
ment, the fundamental component of volt- 
age in a 1JX40 microsecond wave 4 is of 
the order of 10 5 cycles per second. Hence 
components as high in frequency as 10 7 
cycles per second are of significance in 
fixing the wave form. The cathode ray 
oscillograph used to delineate such surge 
wave forms in conjunction with a potential- 
dividing resistor is generally located at some 
distance from the resistor and main dis- 
charge circuit in order to avoid induction 
effects from the large surge currents. Fig. 5 
indicates the common method used for 
connecting the resistor tap point through a 
coaxial cable to the deflecting plates of the 
oscillograph. The resistor is composed of a 
series of resistance cards. Each card has two 
similar windings wound in opposite direc- 
tions and connected in parallel to reduce 

4 I.e., one that rises to crest in If micro- 
seconds and falls again to one-half crest value in 
40 microseconds. 

inductance. The cable connection to the 
cathode ray oscillograph has a surge im- 
pedance usually of the order of 50 ohms, 
whereas the full-scale sensitivity of the 
cathode ray oscillograph is of the order of 
2,000 volts for a cold cathode-type and 200 
volts for a hot cathode-type cathode-ray 
oscillograph. Thus for full-scale deflection 
the currents to a cable considered as a 
resistance would be of the order of 40 or 
4 amperes, respectively. 

Resistance dividers for high-voltage 
surge measurements are not provided with 
corona shields because they would add 
excessive stray capacitance but are often 
so located adjacent to the surge generator 
itself that some degree of shielding and 
field grading along the divider is secured. 
Since a resistor for a 2,000,000-volt surge 
measurement must be at least 700 cm. long 
to provide adequate longitudinal insulations 
and avoid flash over troubles in air, its 
distributed and stray capacitances intro- 
duce disturbing effects unless the resistance 
is held proportionately low. Thus a divider 
resistor for a 2,000,000-volt surge may have 
a resistance as low as 5,000 ohms and, if 
wire-wound, the wire must be of sufficient 
diameter to carry high momentary currents 
without damage. As a rule, accuracies of 
the order of one percent are all that are 
required. If it is desired to study surge com- 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


ponents of the order of 10 8 or 10 9 cycles per 
second the residual inductance of each ele- 
ment and the distributed and particularly 
the non-uniform stray capacitance effects, 
assume especial significance in surge resis- 
tor dividers. For those frequencies, sup- 
plementary capacitance elements (14) may 
be added to the resistance elements of the 
divider in such a manner as to make uniform 
the capacitance shunting effect on each 
element and thus insure proper division of 
the surge voltage and its correct delineation 
by the oscillograph. The present interest in 
better voltage measurements of surge wave- 
fronts of duration less than one microsecond 
should lead to improved designs of surge- 
voltage resistor dividers. 

The cathode ray oscillograph plays such 
an important role in the field of surge- 
voltage (and surge-current) measurements 
that its contribution should be mentioned 
at this point. Although its value for use at 
lower frequencies should not be underesti- 
mated, it stands alone in versatility for use 
in high-frequency, surge, and pulse meas- 
urements. It owes this versatility to the 
small inertia of its electron beam, which 
serves at the same time as the moving ele- 
ment and pointer. The cathode-ray oscillo- 
graph, with a suitable sweep circuit, am- 
plifiers, shunts, and potential dividers, has 
been developed into one of the most useful 
pieces of electrical laboratory equipment 
and covers a range from a few cycles per 
second to frequencies of millions of cycles 
per second. Developments in electron optics 
and of new phosphors are constantly en- 
larging the place of the cathode-ray oscillo- 
graph as a useful precision device in spite 
of the complication of its accessory equip- 
ment. In the future an order of accuracy 
better than 1 percent may be expected. 
For most measurements, and especially 
those of high voltage surges, its high effec- 
tive impedance results from the low capaci- 
tance between its deflection plates. The 
energy loss, arising largely from stray ions 
and electrons within the tube, is so small 
as to be of little concern. Thus in the visual 
or photographic delineation of wave, surge, 
and pulse shape the starting point now 
seems to be a calibrated cathode-ray oscillo- 
graph except in very special instances in the 

low frequency range where a higher order 
of accuracy is required. 

Capacitor dividers. — Because residual in- 
ductance and residual capacitance effects 
in a resistance divider are unavoidable, not 
only in the resistance elements themselves 
but also in their electrical connections, it is 
natural to turn to the use of capacitance 
elements for alternating voltage division. 
The capacitance divider generally consists 
of a single high-voltage guarded-electrode 
capacitor in series with a low-voltage capac- 
itor of very much higher capacitance. The 

linned copper 

Fig. 6. — Compressed-gas capacitor. Diagram of 
Bousman and Ten Broeck capacitor altered to 
include complete shielding of the working elec- 



residual inductance effects in capacitors are 
minimized by properly shaping their sizable 
conducting parts. The residual resistance 
effects are minimized (1) by using high 
grade solid insulation for the electrode sup- 
ports in order to keep electrical leakage 
through and across the surface of the insu- 
lation low; (2) by using free-air or com- 
pressed gases (15, 16) as the dielectric 
because of their low dielectric loss; and 
(3) by using electrodes with well-rounded 
contours and smooth surfaces to insure 
freedom from corona discharge. It is cus- 
tomary to support the guard electrode on 
its own insulation opposite the high-voltage 
electrode and in turn to support the guarded 
low-voltage electrode on the guard. The 
insulation between the guard and the 
guarded electrode is arranged to lie outside 
the high electrostatic field so that dielectric 
losses as well as surface and volume leak- 
age to the working capacitance are mini- 

The range of various types in free air is 
limited by the breakdown voltage of air to 
a practical value of about 15,000 v/cm 
gradient. Churcher (17) has described a 
capacitor for use at 300 kv, cylindrical in 
form, with an over-all diameter of 2 meters 
and a height of 7 meters. This unit is to 
some extent an absolute standard as it 
permits computation of its capacitance 
from dimensions measured under normal 
working conditions. To that extent it is 
preferable to a smaller compressed gas 
type (16) shown in Fig. 6 5 for use up to 300 
kv, having an outside diameter of less than 
1 meter and height over the bushing of 
3 meters, and a shielded capacitance to the 
high-voltage electrode of 50 micromicro- 
farads. However, when the space available 
is limited, the compressed-gas equipment 
may be built to occupy about one-fourth 
the space of a free air unit of the same rating. 
The one advantage, which may be in part 
psychological, of constructing a capacitor 
whose capacitance can be calculated from 
its dimensions lies in the care and precision 
demanded in its design and manufacture. 

5 In this figure the original design of Bousman 
and Ten Broeck has been modified to show com- 
plete shielding of the outer guarded cylindrical 
section by the guard. 

A rather recent and desirable circuit ar- 
rangement using a compressed-gas shielded 
capacitor for testing potential transformers 
is described by Bousman and Ten Broeck 
(16). Their circuit arrangement is similar to 
a Schering bridge, i.e., a bridge in which 
the A and X arms are capacitances instead 
of resistances. The difference lies primarily 
in the mode of supplying the voltage to the 
bridge, the high-voltage arm (A) being 
supplied by connection to the ungrounded 
end of the high-voltage winding of the 
transformer, and the low-voltage arm (X) 
being connected to the ungrounded end of 
the low- voltage winding. Thus if this bridge 
is balanced when supplying power to one 
winding of the transformer, one may de- 
termine from the settings of the bridge arms 
both the voltage ratio and phase angle of 
the transformer. A simple reconnection of 
the bridge arms to a suitable supply per- 
mits quickly checking the constancy of the 
bridge arm components before and after 
ratio and phase angle measurements. Thus 
in this bridge reliance for the ratio measure- 
ment is placed primarily not on capacitanc- 
es but on the . constancy of resistance 
coils which are more suitable as reference 
standards because of their stability, while 
the phase angles are based on the air ca- 
pacitors. An extension in range to higher 
voltages might logically follow the arrange- 
ment of Bousman and Ten Broeck without 
serious reduction in the accuracy of 0.1 
percent claimed for their equipment. 

The subject of high-voltage wave form 
should also be considered in connection 
with capacitance dividers. Offhand, capaci- 
tance dividers would appear to be ideal for 
use with the cathode-ray oscillograph, as 
its impedance is essentially capacitive reac- 
tance. In surge measurements, however, 
it is usually both desirable and convenient 
to locate the cathode-ray oscillograph 
at some distance from the surge circuit 
in order to minimize inductive effects. 
This involves the use of a fairly long 
high-quality cable (preferably coaxial) con- 
necting the divider to the oscillograph so 
that the surge impedance of the cable, 
primarily resistive, rather than the capaci- 
tive impedance of the oscillograph plates 
assumes the major role in the measurement 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


circuit. Thus, as suggested earlier, a surge 
resistor is to be preferred although surge 
capacitor dividers (18) have been used. For 
lower-frequency measurements this objec- 
tion does not exist to the same extent and 
capacitor dividers with cathode-ray oscillo- 
graphs as well as with amplifiers supplying 
electromagnetic oscillographs (string or 
loop in use up to several thousand cycles 
per second) have been found useful in 
delineating wave form with an accuracy of 
a few percent. 

A capacitance-divider method yielding a 
high order of precision in delineating low- 
frequency wave form is due to Silsbee (19) 
who, in line with the early work of Rosa, 
used a point-by-point method. In his device 
a potentiometric balance by means of a 
quadrant electrometer is obtained across 
the low voltage portion of a capacitance po- 
tential divider for as many points in a 
repeated voltage wave as may be desired, 
thus permitting the evaluation of crest-, 
average-, and effective-values of the wave 
form. This method requires a synchro- 
nously driven contactor which may be. set 
accurately for each balance point selected 
for delineating the wave. The accuracy of 
measurement is very high and appears to be 
limited primarily by the steadiness of the 
alternating voltage source, say to a few 
parts in ten thousand. 


Voltage transformer. — The method of 
measuring high alternating voltage in com- 
mon use in the United States employs a 
step-down transformer termed a voltage or 
"potential" transformer by the manufac- 
turer in order to designate its intended use 
in voltage and power measurement. The 
high-voltage winding is connected across 
the terminals of the voltage source to be 
measured, a voltmeter is connected across 
the low-voltage winding and its reading is 
multiplied by the ratio of transformation to 
obtain the value of the high voltage. The 
ratio of voltage of the high-voltage winding 
to that of the low-voltage winding of a 
well-designed transformer remains nearly 
constant over a wide range of voltage. The 
measured values of ratio of well-built 
transformers kept under normal laboratory 

conditions have been found to remain re- 
markably constant over long periods of 
time, variations in ratio requiring for their 
detection measurements having an accuracy 
of the order of 0.01 percent. Departures 
from nominal ratio have been determined 
by resistance bridge methods (20) up to 30 
kv with a precision of 0.01 percent and by 
capacitance bridge methods up to 132 kv 
with a precision of 0.1 percent or better 

Where symmetrical multiple high- voltage 
windings are provided in a voltage trans- 
former the following "series-parallel prin- 
ciple' ' has been found valid for extrapolat- 
ing measurements of ratio factor at low 
voltage to permit their use at higher volt- 
age. Within the voltage limits of a high- 
voltage shielded resistor or capacitor, and 
with the high- volt age windings in parallel, 
measurements of ratio factor are made to 
cover the voltage per coil range of the low- 
voltage winding. The nominal ratio for 
series connection is then multiplied by the 
measured ratio factor at the same volts per 
coil for the parallel connection. Reliable 
measurements of effective alternating volt- 
age by this method can be made to better 
than 0.1 percent (20) and are in use up to 
250 kv. The cost of transformers with such 
symmetrical multiple windings increases 
rapidly with voltage. Other less expensive 
and less accurate (1 percent to 0.5 percent) 
transforming devices such as (1) a number 
of small chain-connected or cascaded trans- 
formers (21, 22) and (2) a high-voltage 
resistor (23), reactor (24), or capacitor 
(25) in series with a small transformer, have 
been introduced in Europe for measuring 
high voltage but have not met with general 
favor in the United States. 

Supply transformer with high-voltage wind- 
ing tap or with voltmeter coil. — Although the 
use of a voltage transformer with only an 
instrument connected as the burden on the 
low-voltage winding represents the ideal 
arrangement, it should be mentioned that 
adequately precise values of high voltage 
may often be deduced from voltage meas- 
urements made on the low-voltage input 
windings. This is particularly true if the 
resistance and leakage reactance of the 
high-voltage winding are low and the cur- 



rent drawn by the load on the high-voltage 
winding is small. Better still, an instrument 
connected between ground and a tap point 
on the high-voltage winding (26) near its 
grounded end may serve for measuring 
a relatively low voltage which will be pro- 
portional to the high voltage except in so 
far as the voltage across the tapped section 
of winding is affected by distributed and 
stray capacitance currents which flow 
through it from the rest of the high-voltage 
winding. Fig. 7 shows three such 350 kv 
60 c/s 1,000 kva transformers in the High 
Voltage Laboratory of the National Bureau 
of Standards. These units are shown con- 
nected in cascade to give 1,000,000 volts. 
Each unit has a high- voltage winding tap 

connection giving an accuracy of voltage 
measurement of about 2 percent at full 
load and correspondingly better accuracy 
at lighter loads. Still another arrangement 
consists in providing a third winding or 
voltmeter coil (27) so located relative to the 
low- and high-voltage windings that the 
magnetic flux linked by it automatically 
takes into account any voltage drop in the 
high voltage winding arising from the load 
connected to it. An accuracy of one-half 
of 1 percent at full-load leading current and 
better accuracy at smaller loads is claimed 
for a good design. These latter arrange- 
ments are not considered to be as trust- 
worthy as the use of a separate voltage 

Fig. 7. — Three transformers connected in cascade to give 1,000,000 volts. 

Feb. 15, 1948 defandorf: the measurement of high voltage 


Fig. 8. — Bipolar generating voltmeter. 


The generating (rotary) voltmeter (28) 
first described by Kirkpatrick has been 
useful in estimating local potential gradi- 
ents and as an auxiliary control device. It 
has been likened to a d-c generator, but 
instead of currents induced by moving 
wires in a magnetic field it employs a 
configuration of electrodes which permits 
the commutation of charges induced on 
plates alternately exposed to and shielded 
from an electrostatic field. Fig. 8 shows the 
diagram of a bipolar generating voltmeter. 
Fig. 9 shows a ceiling-mounted type de- 
signed by Behr that employs a sector disk 
as the rotating element. 

The fundamental idea has been incor- 
porated in many designs. Because of the 
necessity of alternately exposing and shield- 
ing the active plates (electrodes) the gen- 

erating voltmeter appears to offer greater 
promise for use in relative measurements 
than in absolute measurements. Thus it is 
usually first calibrated in terms of other 
satisfactory low-voltage standards, prior 
to use at higher voltage, but it may be used 
in those cases not requiring a high order of 
accuracy and thus amenable to simplifying 
assumptions for the purpose of computing 
the high voltage from dimensions. By so 
shaping the high voltage electrodes that 
discharges and consequent space charges 
are avoided it may be used by extrapolation 
techniques to extremely high voltages both 
on direct and alternating voltage. Without 
an incorporated high-voltage electrode it is 
essentially a gradient measuring device. It 
has been useful in a study of atmospheric 
electric charge and field phenomena re- 
sponsible for lightning (29), and as a voltage 

Fig. 9. — Rotating segments of Behr generating 




measuring device associated with high- 
voltage generators of the Van de GraafT 

When the generating voltmeter is used 
as a voltage measuring component (i.e., as 
an instrument for determining a value of 
gradient to be multiplied by a constant 
factor in order to obtain the total voltage 
difference) it is not sufficient that the fields 
in the immediate neighborhood of the in- 
strument shall be below corona or dis- 
charge-forming values, but it is equally 
necessary that the gradients at surfaces in 
the neighborhood of the sample field as 
well as those at the high-voltage electrode 
be below discharge-forming values and that 
there be no nearby relatively large sources 
of air ionization or dust. The effect of dust 
and charged particles may indirectly pro- 
duce serious distortions in high fields by 
being deposited and aligned on otherwise 
smooth polished metal surfaces so as to 
form the necessary points for initiating 
discharges and consequent space charges 
that would otherwise not exist. If the 
charge on the surface of the segments al- 
ternately exposed to the electric field is to 
be in definite ratio to the total field, then 
at no place in the field may the gradient 
exceed the approximate breakdown gradi- 
ent, 30 kv crest/cm, in air at NPT. The 
practical limit is actually much lower. For 
instance, the value of average gradient at 
which self-propagating discharges occur 
in the case of thunderstorms appears to be 
about 10 kv/cm. In the case of the Brooks 
electrometer (30) discharges have been 
known to occur between the parallel faces 
of the electrometer plates when the voltage 
applied divided by the spacing had a value 
of 5 kv eff/cm. Thus in air at NPT it is 
doubtful that for a generating voltmeter an 
average value of gradient exceeding 7 kv 
max/cm with an upper limit of 20 kv/cm at 
the electrode surfaces can be employed con- 
tinuously without some error in measure- 
ment arising from the above causes. Pre- 
sumably this practical limit arises from 
the almost ever-present small particles of 
dust (insulating or semi-conducting) which 
when deposited on the surface upset the 
ideal condition of pointless smooth surfaces 
one wishes to assume after the instrument 

maker has done his best to produce them. 
With reasonable care and cleanliness in 
assembly, dust has not been found to be a 
problem in generating voltmeters operating 
within a pressure chamber at higher gradi- 
ents (31). 

When used with suitable precautions, the 
generating voltmeter offers a means of ob- 
taining information not readily available in 
other ways. For instance when its corn- 
mutating device is provided with a phase- 
shifting mechanism and the rotor is syn- 
chronously driven it provides a means of 
determining wave form (32) at high voltage 
with practically no load on the source being 
investigated. It is effective on a rippled 
direct voltage as well as on alternating 
voltage. WTien used for determining wave 
form it may be located in an undisturbed 
and readily accessible portion of the high 
voltage field. The generating voltmeter 
method of determining wave form should 
be good to a few tenths of 1 percent. 


Sphere and rod gaps. — If an accuracy of 
the order of 3 percent in determining crest 
voltage is sufficient, as in the case of insu- 
lator testing, the sphere spark gap in air is 
useful as a crest voltage measuring device. 
For voltages above 17 kv spheres of 6.25 
cm diameter and larger (up to 2 meter 
diameter for approximately 2,000,000 volts) 
operated under controlled conditions serve 
as voltage standards for electrical break- 
down measurements of dielectrics (33). 
Earlier theories of breakdown of sphere 
gaps assumed that pure air, as well as other 
gases, has a definite breakdown strength or 
breakdown gradient at normal pressure and 
temperature. Paschen's law relating the 
length of the breakdown gap with air den- 
sity permits correction for usual tempera- 
ture and pressure variations. Russel, Dean, 
Peek, and others have given empirical re- 
lationships based in part on electrostatic 
field theory for both sphere and cylindrical 
gaps, which, although they fit well in a 
limited domain of pressure and tempera- 
ture, are not so satisfying as relations (34) 
developed at a later date on the basis of the 
newer atom-physical background. The de- 
tailed work on discharges in gases by Loeb 

Feb. 15, 1948 

defaxdorf: the measurement of high voltage 


(85) and his collaborators as well as much 
other valuable work in this field has pro- 
vided a fairly satisfactory explanation of 
some of the scattering of breakdown values 
of spark gaps. Meek (36) suggests that in 
the case of the shorter gaps the electron 
avalanche initiated at the cathode con- 
stitutes the usual initial process whereas in 
the case of larger gaps the mechanism is 
more akin to the positive streamer dis- 
charge of lightning and originates within 
the gap at some distance from the electrode. 
For gaps of intermediate length there lies a 
domain in which the initiation may be of 
either variety and such gaps show a larger 
scattering of breakdown values. The prob- 
ability of a free electron existing in the right 
location to initiate a discharge is, of course, 
an important factor in the scattering of the 
initial breakdown voltage when the time 
of application of voltage is extremely short. 
Irradiation (37) of small electrodes of the 
gap with radium or with ultraviolet light to 
produce photoelectrons is helpful in reduc- 
ing this type of scattering but appears not 
to be necessary in the case of large spheres 
because of the much higher probability, in 
the larger volume of air between the spheres 
of the presence of an initiating ion or elec- 

An idea of the effectiveness of ultraviolet 
irradiation in reducing scattering may be 
cited in the case of 12.5 cm diameter brass 
spheres. Irradiation of the spheres by an 
open carbon arc reduced the scattering of 
individual 60-cycle sparkover values by a 
factor of 5 as compared with the results 
obtained without irradiation. On the other 
hand the average values of sparkover volt- 
age were lower by from two to five percent 
in the irradiated case, the amount of lower- 
ing being dependent on the intensity of 

Basing his work on the detailed informa- 
tion now available on the mechanism of 
spark formation Ver Planck (38) appears to 
have successfully correlated the enormous 
amount of data on sphere spark gaps. 

From the very nature of the spark be- 
tween spheres and of corona on cylinders, 
voltage measurements based on sphere 
gaps and corona cylinders (34), because of 
their dependence on surface shape, cleanli- 

ness of the surface, and cleanliness of the 
air, appear to be limited to a rather low and 
questionable accuracy, say from 1 to 3 
percent. When extreme precautions are 
used by following a ritual of cleansing, the 
use of ultraviolet irradiation of the spheres, 
limitation by resistance of surface pitting 
by the spark current, and by insuring 
"cleanup" of the sparking surfaces of the 
sphere gap through preliminary sparking, a 
series of 10 or more sparkover values often 
may be observed to agree to within ±0.1 
percent. However, this apparent high pre- 
cision, equal to that of the high-grade indi- 
cating voltmeter employed as part of the 
equipment in making such observations, is 
deceptive. Painstaking observations made 
the following day under seemingly identical 
conditions may agree among themselves to 
the same precision but their average value 
will almost invariably differ by several 
tenths of one percent, and sometimes by 
more than 1 percent from those made on the 
previous day. In spite of its low order of 
accuracy the sphere gap serves as a com- 
mercial standard (33) for high alternating- 
and surge-voltage measurements appar- 
ently because of its basic simplicity. In 
larger sizes it gives a useful measurement of 
the maximum value of surge voltage and 
essentially serves as a voltage limiter when 
it is used in parallel with a device under- 
going voltage-withstand tests. 

Simple needle gaps, as a matter of his- 
torical interest, were once accepted as al- 
ternative standard voltage measuring gaps, 
but as a result of the inherently high scat- 
tering of values of sparkover voltage, varia- 
tion in sparkover values with changes in 
humidity, and large scattering in sparkover 
voltage when used for measuring surge 
voltage, they have been discarded as stand- 
ards. Rod gaps (39) in which the elec- 
trodes are used repeatedly as contrasted 
with the formerly prescribed use of new 
#00 needles for each sparkover of the needle 
gap, have sparkover values which are also 
affected by humidity to about the same 
extent as the flashover of porcelain insula- 
tion and appear to have replaced needle 
(or point) gaps for those uses where gaps 
are desired for correlating the flashover 
voltage of insulators. 




The Ionic Wind Voltmeter described by 
Thornton, Waters, and Thompson (40) 
merits some mention because it represents 
a unique application of a thermal method 
to the measurement of high alternating 
voltage. Use is made of the cooling effect 
on a heated filament arising from "elec- 
tric wind." (See Fig. 10.) The heated fila- 
ment, with a suitable grounded shield, is 
located at a distance from the high-voltage 
electrode in such a position that, although 
in the electric field, it will always be below 
corona-forming voltage. Ions that may be 
present move back and forth as a result of 
the alternating electric field and in striking 
neutral molecules increase the general 
molecular motion. This results in an in- 
crease in cooling effect on the heated fila- 
ment proportional to the electric field. A 
filament that has a high temperature co- 
efficient of resistance is connected in one 
arm of a Wheatstone bridge. The bridge 
out-of-balance indicator is then calibrated 
in terms of the high voltage applied to the 
Ionic Wind Voltmeter. Although this de- 
vice may be constructed to have good 
sensitivity and is useful as a control device 
or relay, its indications are affected by 
change in wave form and an accuracy of 
only + 2 percent is claimed for it. 


Electrostatic voltmeters and electrom- 
eters basically depend for their indication 
on a measurement of the force of attraction 
between charges on the movable portion 
of one electrode surface and charges of 
opposite sign on another fixed electrode 
surface. By arranging the movable portion 
of the electrode to be part of a suitable 
geometric surface — sphere, ellipsoid, or 
plane — it is possible to devise an instrument 
in which the voltage applied can be com- 
puted theoretically from measured dimen- 
sions and the measured force of attraction. 
An electrometer designed to approximate 
quite closely the theoretical assumptions as 
to the conductor shape and relative di- 
mensions required for simple theoretical 
computations, and thus to permit computa- 
tion of the value of applied voltage from 

dimensions and the resulting force, is 
termed an absolute electrometer. This is in 
contradistinction to the term electrostatic 
voltmeter, which signifies an instrument 
that may be used for relative measurements 
but that requires calibration by means of 
some other standard of voltage measure- 
ment. A number of designs of high-voltage 
electrostatic voltmeters (41-51) have been 
constructed. Such instruments require 
much less electrical energy for their opera- 
tion than an electrodynamic instrument 
with a series resistor. Corners rounded suf- 
ficiently to avoid electric discharges, high- 
quality insulation, and electrode spacing 
adequate to prevent discharges are pre- 
requisites in the construction of both elec- 
trostatic voltmeters and electrometers. 

Electrostatic voltmeters. — With a few ex- 
ceptions (46, 47, 4$) the high voltage elec- 
trostatic voltmeters follow the pattern of 
the Kelvin (52) guard-ring electrometer by 
having a guard ring for the attracted disk 
or movable electrode while placing less 
emphasis on the flatness of the movable 
electrode. Provision is usually made for 


'I ' l 








Fig. 10. — Ionic wind voltmeter. 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


change in range by adjusting the spacing 
between the high-voltage electrode and 
grounded electrode of which the moving 
element usually forms a part. It is interest- 
ing to consider the developments in these 
instruments over a few years as shown in 
Figs. 11, 12, and 13. In the early design of 
Abraham and Villard (41) (1911), Fig. 11, 
the curvature of the disk and guard is 

Fig. 11. 

-Abraham and Villard electrostatic 

nearly spherical. The moving system is 
rather massive with a separate damping 
chamber within the guard. The Imhof (44) 
(1926) design, Fig. 12, provides a flattened 
guard ring and fits the light-weight at- 
tracted electrode into its own damping 
chamber. Starke and Schroeder (43) (1928) 
in one model, Fig. 13, employed a relatively 
larger flat-guarded electrode and a flat- 
strip suspension for the movable rectangu- 
lar flat electrode, P, providing it with a 
mirror instead of a mechanical pointer, as 
well as with a balanced damping chamber 
arrangement. Nearby objects would be 
increasingly less effective in producing de- 
flection errors at equivalent spacings in the 
later voltmeters. These voltmeters may be 
read to closer than 1 percent but unless 
calibrated in place, especially when used at 
large spacings at maximum rated voltage, 
are likely to be affected by nearby objects. 

Ellipsoidal voltmeter. — The ellipsoidal 
voltmeter of Thornton and Thompson 
(58), illustrated in Fig. 14, is of the nature 
of an electrometer and satisfactory theoret- 
ical equations have been developed for it. 
It, like the electrometer, depends on a 
relatively undistorted axial field if it is to 
be used as an absolute instrument. 

The moving element consists of a metallic 
ellipsoid of revolution carried on a bifilar 
silk suspension and is provided at its lower 
end with a reflecting mirror and damping 

Fig. 12. — Imhof electrostatic voltmeter. 



Vy ;"'V.v •>:.•• 

Fig. 13. — Starke and Schroeder electrostatic voltmeter. 

vane. The ellipsoid is suspended with its 
long axis horizontal and is centrally located 
between two vertical circular plates facing 
each other and at an adjustable distance 
apart. Polarization of the ellipsoid by the 
electric field established between the plates 
by the source of voltage to be measured, 
gives rise to a mechanical couple tending to 
align the ellipsoid, which is initially set at 
an angle to the horizontal axis perpendicu- 
lar to both plates. Although the change in 
the angular deflection of the ellipsoid may 
be used as a measure of the voltage applied, 
a more sensitive and rather interesting 
method of voltage indication has been con- 
trived which involves measuring the in- 
crease in frequency of its swings when volt- 
age is applied. Thus: 

E = k(n 2 -n 2 ) 



where E is the electric field strength, A; is a 
constant found from dimensions, n is the 
number of swings per second with the volt- 
age on, and n is the number of swings per 
second with the voltage off. 

Sufficiently large plates are used and at 
such a separation that the electric gradient 
E throughout a considerable volume along 
the axis would be substantially uniform if 
the ellipsoid were not present. Since the 
disturbance effected by the ellipsoid is 
small, the voltage applied to the plates is 

V = Ed, 
where d is the plate separation. 


The instrument described was designed 
for measurements up to 200 kv. As shown 
schematically in Fig. 14 the plates were 
140 cm in diameter and were spaced as 
much as 100 cm apart (i.e., average gradi- 
ent of 2 kv per cm). An exploration of the 
field indicated that it was quite uniform at 
the midpoint between the plates within a 
radial distance of 25 cm; however, no study 
of its longitudinal variation appears to have 
been made. Great care was taken to insure 
that the ellipsoids of revolution were ac- 
curately shaped so that the theoretical 
relationships would apply. One ellipsoid 
of duralumin was 3.9692 cm long and 0.5970 
cm in diameter and weighed 2.0496 grams. 

The electric gradients at the tips of the 
ellipsoids are considerably higher than the 
average gradient between the plates. These 
gradients must be kept well below corona- 
forming values if disturbing effects from 
electric wind are to be avoided. In its prac- 
tical form where it is to be used as a de- 
flection instrument for laboratory or shop 
measurements an insulating enclosure is 
provided for the suspension, ellipsoid, mir- 
ror, and damping mechanism to shield 
them against wind and dust. At the bottom 
end of the moving-system assembly the 
damping vane is suspended in a damping 
chamber attached to a tube whose upper 
end is cemented to the bottom of a hollow 
glass sphere in which the ellipsoid is cen- 
tered. The upper end of the sphere is ce- 
mented to a second tube housing the bifilar 

Feb. 15, 1948 



suspension and an adjustable suspension 
control for changing the period of swing. 
The control mechanism and the torsion 
head are mounted at the supported upper 
end of the tube. A spherical shape was used 
for the hollow glass sphere enclosure for the 
ellipsoid so as to permit a theoretical eval- 
uation of the effect of the dielectric of the 
sphere on the electric field E at the ellip- 
soid. The theoretical correction derived 
for this spherical glass enclosure agreed 
well with experimental results when the 
relative humidity was not high enough to 
cause electrical surface leakage. 

The ellipsoidal voltmeter is said to be 
accurate to 0.1 percent and to be only 
slightly affected by humidity. Because of 
low average gradient, 2 kv eff/cm, it ap- 
pears to be more bulky than other electro- 
static voltmeters and is of interest mainly 
because it is a unique arrangement per- 
mitting absolute measurements. 

Sparkless sphere-gap voltmeter. — Large 
spheres ordinarily used as sphere spark-gap 
voltmeters in measuring high crest-voltage 
have been modified to permit their use as 
electrostatic voltmeters for measuring the 
effective value of voltage at spacings 
slightly in excess of sparking distances. 
This arrangement has been called a spark- 
less sphere-gap voltmeter (47). 

Hueter (46) employed a vertical ar- 
rangement of 1-meter spheres. The upper 
high-voltage sphere was supported on a 
spring whose additional extension as a 
result of the electrostatic force was magni- 
fied by a lamp-mirror-scale arrangement. 
The spring and the mirror optical-lever 
arrangement were mounted within the 
sphere shank which was provided with a 
small window. An external arc-lamp and 
scale were mounted on an adjacent wall 
in the laboratory and gave satisfactory 
readings in daylight. The vertically adjust- 
able lower sphere was grounded and its 
driving screw mechanism was arranged to 
indicate the gap length. The weight of the 
upper one-meter sphere was 60 kg and for a 
75 cm gap the electrostatic attraction was 
approximately 800 grams at 1,000,000 
volts. In order to minimize effects of 
changing gap length, only small displace- 
ments (less than 0.5 percent of the gap 
length) of the spring-suspended sphere were 

used. An oil-cup damper made the sphere 
motion nearly aperiodic. An accuracy of 1 
percent was claimed. 

Sorensen (47, 4&) employed a horizontal 
arrangement of 1-meter spheres with rather 
long slender shanks presumably in order 
to reduce effects of attraction arising from 
the shanks. The electrostatic attraction of 
the grounded sphere could be readily meas- 
ured as it was supported by suspending its 
shank by four ropes tied to the shank at 
the apices of the two thus-formed F-sus- 
pensions. The upper ends of the ropes were 
attached to ceiling members. This laterally 
stable suspension possessed only a small 
longitudinal stability so that differences in 
longitudinal electrostatic forces of several 
hundred grams could be measured to better 
than one gram. A small wire in line with and 
attached to the end of the grounded sphere 
shaft ran over the rim of a bicycle wheel 
thus insuring low friction. A small weight 
pan attached to the end of the wire per- 
mitted weighing the force of attraction. 

These two sphere-electrometer devices 
represent useful laboratory tools for they 
can be calibrated and used as voltmeters 
for measuring effective voltage. They can 

Suspension adjustment 

Fig. 14. — Ellipsoidal voltmeter. 



also be employed in the usual manner as 
sphere spark-gaps for the measurement of 
crest voltage. If one is satisfied with the 
accuracy of the sphere spark-gap as a 
high-voltage standard, measurement of the 
voltage wave form permits calibration of 
these devices as electrometers for the meas- 
urement of effective voltage without re- 
course to other methods. To a very limited 
extent they may be considered to be abso- 
lute electrometers, particularly if adequate 
relative clearances to the floor, walls, leads 
and other conductors and insulators are 
maintained so that approximate corrections 
can be made for the presence of these ob- 
jects. Simple theoretical calculations of the 
force of attraction can be made by the 
method of image charges in the case of two 
insulated or one grounded and one insulated 
sphere. For a 25-cm spacing of his 1-meter 
spheres Sorensen considered that no cor- 
rection was necessary, and at 35 cm the 
effect of shanks and other nearby objects 
was said not to exceed 1.5 percent. 

One may hope for an accuracy approach- 
ing 0.5 percent with the design of Hueter 
in making relative measurements up to one 
million volts. This would involve ample 
clearance for the spheres and first calibrat- 
ing by an accurate lower voltage method 
using a separation of the spheres adequate 
for one million volts and making certain 
that corona-free electric field conditions 
exist in the neighborhood. 

Disk electrometers. — About 1880 Lord 
Kelvin made an outstanding contribution 
by incorporating a guard ring for the disk 
of the attracted disk electrometer. The 
guard ring not only validified the use of the 
simplifying mathematical assumptions in 
computing the electrostatic force of attrac- 
tion of the disk but it established the basis 
for a design whose readings were less ef- 
fected by nearby objects. When coplanar 
with the guard ring the force 

8d 2 


where V s the voltage is in electrostatic 
units, A is the area of the disk, and d the sep- 
aration from the opposite grounded plate. 
For deflections of a guarded disk away 

from coplanarity, Snow (54) has developed 
a mathematical solution which takes into 
account the change in force arising from 
change in position of the disk. This change 
in force might at first appear only to vary 
inversely with the separation d (shown by 
a simple differentiation of equation (3)). 
However, Snow has evaluated the addi- 
tional change in force arising from the 
redistribution of the charges, which in the 
case of a protruding disk results in an 
increase in concentration of charge at the 
edge of the disk at the expense of the charge 
on the adjacent edge of the guard plate 
and vice versa in the case of a retracted 
disk. Troublesome instability in disk elec- 
trometers arising from this latter com- 
ponent of force has for a long time been 
recognized as a weakness in electrometers 
designed to cover a wide range of voltage 
measurement by adjusting the spacing d. 
One may either elect to provide a linear 
restoring force adequate for all spacings d 
at a considerable sacrifice in sensitivity at 
large spacings or provide for some adjust- 
ment of the restoring force with change in 
spacings at nearly constant maximum al- 
lowable gradient. The restoring force re- 
quired to balance the electrostatic attrac- 
tion in high voltage electrostatic volt- 
meters and electrometers has been pro- 
vided by the following devices : 

(1) Suspension of the moving electrode on one 
arm of a gravity balance, i.e., change of restoring 
force secured by adjustment of e.g. of balance 
relative to central knife edges (52, 30). 

(2) Suspension of moving electrode on a coiled 
spring (46, 61). 

(3) Pendulous suspension of electrode (49). 

(4) Suspension of disk electrode assembly on a 
flat-strip (torsion) suspension (43, 55). 

(5) Suspension of the disk from a metallic 
membrane (50, 51). 

(6) By combining the torque provided by a 
flat-strip supporting suspension with that pro- 
duced in a current-balance arrangement of coils 
which permits adjustment of the restoring force 
(55) by change in current. 

The last three types merit special men- 
tion because of the novelty of their ar- 
rangement. Fig. 13 illustrates the flat-strip 
suspension arrangement of Starke and 
Schroeder. The sixth of the above arrange- 
ments, used by Nacken (55), employs two 
pairs of "current balance" coils in a con- 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


nection that permits adjustment of sta- 
bility as well as of sensitivity. The restoring 
force arises in part from the vertical strip 
suspension which carries the disk with its 
plane vertical as well as the two similarly 
mounted astatically connected current bal- 
ance coils. One pair of field coils with cur- 
rent if is connected in series aiding one 
moving coil and provides a torque propor- 
tional to if for balancing the electrostatic 
force on the disk; the other pair of coils 
with current i m is connected in series op- 
posing so that by reversing i m and adjusting 
its value relative to the current in the mov- 
ing coil the restoring force of the strip 
suspension toward the null or coplanar 
position of the disk may be either opposed 
or aided thus altering the stability of this 
system by the simple adjustment of the 
current in this pair of coils. Thus this ar- 
rangement permits a desirable adjustment 
toward higher sensitivity at large spacings 
of the electrodes where the deflecting force 

The fifth of the above arrangements is 
exemplified in the devices of Rogowski and 
Bocker (50, 51) (illustrated in Fig. 15) in 
which an elastic diaphragm D provides the 
restoring force for the attracted disk. The 
diaphragm supports an iron-cored coil 
(above) which is actually part of a current 
transformer ("Messdose") constructed 
with two air gaps in its iron core. The fixed 
coil with its core form the other part of the 
current transformer which is supplied from 
a voltage-regulated alternating-current sup- 
ply. Very small changes in the air gap 
(moving coil position) suffice to produce 
full scale deflections of an ammeter con- 
nected across the moving coil and this am- 
meter is calibrated to read the high voltage. 
As part of this particular moving system, 
arranged for use in a compressed gas en- 
closure, a force coil (Druckspiile) K is sus- 
pended below in an iron-clad solenoid es- 
pecially constructed to permit a measure- 
ment of force in terms of solenoid current. 
This arrangement permits, prior to assem- 
bly of the unit in the compressed gas cham- 
ber, a direct calibration with known weights 
on the disk for current in the solenoid 
against deflections of the diaphragm as in- 
dicated by the ammeter. This preliminary 

calibration with known weights against 
current is made in order to permit subse- 
quent checking of the calibration of the 
electrometer by means of the current in- 
stead of weights after it has been filled with 
compressed gas. 

Brooks absolute electrometer. — The ab- 
solute high voltage electrometer of Brooks 
(30) (Figs. 16 and 17) is illustrative of what 
can be accomplished in the way of precision 
when the attracted disk is supported on 
one arm of a gravity-type balance. Brooks's 
modification of the Kelvin electrostatic 
attracted disk electrometer with guard ring 
was designed for use in free air up to 275 
kv on alternating voltage and was arranged 
to allow a step by step experimental evalua- 
tion of errors not readily calculable. In ad- 
dition to a guard ring, it employed guard 

Fig. 15. — Rogowski and Bocker compressed- 
gas electrostatic voltmeter. 



hoops equally spaced between the guarded equally spaced hoops are maintained at 
and grounded electrodes. These guard equally spaced potentials corresponding to 
hoops, spaced 2 cm apart, are connected to the voltage applied between the high-volt- 
tap points on a potential dividing capacitor age (upper) and grounded (lower) plates of 
also across the voltage source. Thus the the electrometer. The hoops not only screen 

Fig. 16. — Brooks attracted disk electrometer (schematic). 

Feb. 15, 1948 

defandokf: the measurement of high voltage 


the disk from external fields, but by their 
potential distribution tend to correct, 
within their diameter, electric field distor- 
tion that would otherwise arise from edge 
effects at the upper guard ring and lower 
plate. A mathematical solution of the con- 
tribution to force on the disk arising from 
these hoops has been carried through by 
Snow (54) and Silsbee (30) and has been 
demonstrated to be adequate by suitable 
experimental tests. 6 For instance by omit- 
ting hoops, or electrically shorting various 
sets of hoops, their potential distribution 
was changed drastically. For such condi- 
tions when the corresponding corrections 
were applied based on the mathematical 
solution and experimentally measured po- 
tential distribution, very good agreement in 
the measured values of voltages was ob- 

As a result of the work on the Brooks 
Absolute Electrometer, which was com- 
pared with the transformer voltmeter 
method of measuring voltage, it was con- 
cluded that this instrument is reliable for 
absolute determinations to about 0.01 per- 
cent. This work was limited to 100,000 
volts — approximately one-third its rated 
voltage — because the clearances in the 
space in which the equipment was housed 
were inadequate. Results up to full rated 
value, 275 kv, are not yet available. (With 
the cessation of the war it is expected to 
extend the measurements to higher values 
of voltage in the modern High Voltage 
Laboratory of the National Bureau of 

6 The s imple equation (3) may be rearranged 
V„ = 2dV2F/A. This equation may be thought 
of as being satisfactory for measurements yield- 
ing an accuracy of 1 per cent. If an accuracy of 
0.01 per cent is desired, correspondingly refined 
experimental techniques must be employed and 
additional physical measurements must be made 
as indicated by the larger number of terms in the 
complete equation for the voltage in electro- 
magnetic units of the Brooks electrometer, 

_ Mc+h d )\/2M7g r A v -A 
M ~ (n+r d ) L 2M F g 

+17" {S * 
M F 


f(S v -b m ) fkn 

2q 2 


Compressed gas electrometer. — A consider- 
ation of the compressed gas electrometers 
of Tschernyscheff (56), Palm (57), Rogow- 
ski, Bocker (50, 51), and others who at- 
tained accuracies approaching 0.1 percent 
leads one to consider whether additional 
design features might be incorporated in 
the compressed gas type in order to give 
an accuracy approaching that of the Brooks 
electrometer in free air. 

Actually the matter of precise weighing 
of the forces (less than 2.5 grams) on a 
16-cm diameter disk, involved in the Brooks 
electrometer offers no serious problem in 
free air aside from the necessity of providing 
a carefully thermostated enclosure to 
avoid air currents. The possibility of having 
correspondingly higher forces to measure 
has its appeal. In the Rogowski and Bocker 
voltmeter the maximum force of attraction 
of the disk may reach 250 grams and is 
measured to 0.2 grams. Thus the forces 
employed in measurements are one hundred 
times as large as in the Brooks electrom- 
eter, however, the relative accuracy of 
measurement of this larger force is less in 
their device. 

At the outset one would have to provide 
a force-measuring device approaching the 
accuracy and repeatability of a high-grade 
gravity balance, i.e., something better than 
that incorporated in the Rogowski and 
Bocker electrometer. Smaller-scale length 
measurement would have to be made with 
about the same relative accuracy and the 
flat metal surfaces of the plates would re- 
quire a high quality optical finish. The 
greatest loss involved in the use of com- 
pressed gas rather than free air, as in the 
Brooks electrometer, lies in the relative 
difficulty of making and checking mechani- 
cal measurements under pressure before 
and after voltage observations. 

The immediate gain from the use of 
compressed air (15) or other gases (5 IB) 
is to increase sparkover voltages nearly in 
proportion to pressure. For instance in the 
design for use to 400 kv alternating, Bocker 
used carbon dioxide at 15 atmospheres 
pressure which permits a gradient of 100 
kv eff/cm as compared with 2.5 kv eff/cm 
in the Brooks electrometer as limited by the 
present potential dividing capacitor con- 



nected to the hoops. Gaseous "freon" (58) 
(dichlorodifluoromethane) has between 
smooth electrodes about 2.3 times the 
breakdown strength of air at pressures up 
to six atmospheres where it still remains 
gaseous. In the case of points or sharp 
edges the relative breakdown strength of 
freon is still higher. Freon has the disad- 
vantage of breaking down into highly 
corrosive products if corona or other dis- 
charges actually take place in it, making it 
less desirable than carbon dioxide on that 
account. Sulfurhexafluoride (58C) appears 
to offer some advantages over freon be- 
cause of its greater chemical stability and 
higher equilibrium pressure at normal 

The greater size of the "free air" as 
compared with the "compressed gas" elec- 
trometer appears to be its chief drawback, 
making it too cumbersome at the highest 


The cathode ray oscillograph and electro- 
static (and magnetic) analyzers are ex- 
amples of devices that employ the deflection 
of free-moving charged particles. Employ- 
ing for this discussion the relations given 
by Hanson and Benedict (59), if a slowly 
moving stream of charged particles is 
accelerated in vacuo along an electrostatic 
field of total voltage V a then 

so that 

V a e 

-mv l 


where m is the mass of the particle, v its 
velocity, and e the charge on the particle. 
When a stream of particles with the velocity 
v is directed between parallel plates perpen- 
dicular to the electric field established by 
the voltage V d between the plates then for 
the idealized arrangement of plates at the 
far edge the deflection 

d=V d el 2 /2Smv 2 


where I is the length and S is separation of 
the plates. For nonrelativistic velocities if 
L is the distance to and D the deflection at 
the screen or receiver 

V a /V d = Ll/2SD, 


and if the relativistic velocities are taken 
into account (59) 

V a /V d ( 1 )-Ll/2SD, 

\ 2(E +V„e)J 


D = Ld/ (1/2) = V d eLl/Smv 2 


where the rest energy E = m c 2 and m is 
the rest mass and c the velocity of light. 

For relation (8) it may be shown that 
the relativistic correction is about 1 percent 
for electrons accelerated by a voltage V a = 
10.5 kv and increases to about 14 percent 
at 200 kv. The existence of a relativistic 
correction of such magnitudes on account 
of high electron velocity may be looked on 
as somewhat of a nuisance and as a limita- 
tion of the cathode-ray oscillograph when 
used for the direct measurement of high 
voltages. Applications of the cathode-ray 
oscillograph therefore seem to have been 
limited to the measurement of lower volt- 
ages. As is well known, the beam-accelerat- 
ing voltage source for V a is generally main- 
tained as constant as possible in order to 
preserve the sharpness of the cathode spot. 
The voltage to be measured (or a fraction of 
it from the voltage divider) is applied as the 
voltage V d to the deflecting plates. V d is 
kept sufficiently low to leave the deflection 
D materially unaffected by the relativistic 
mass correction for velocity of the electrons. 

At first thought the fundamental simplic- 
ity of this method of measuring voltage is 
decidedly appealing. It applies the accel- 
erating voltage field directly to the elemen- 
tary charge of the electron unhampered by 
additional matter. As indicated above and 
by equation (8) there is a disadvantage 
because the resulting high electron veloci- 
ties even for relatively low-voltage accel- 
erating fields become so large that relativ- 
istic corrections for moving charges must 
be introduced. Thus on second thought, the 
method involving the acceleration of elec- 
trons appears far from ideal but an exami- 
nation of equation (5) indicates the ad- 
vantage to be gained by using charged par- 
ticles of greater mass than the electron be- 
cause of the resulting lower velocity. Al- 
though what one might consider a practical 

Feb. 15, 1948 

defandorf: the measurement of high voltage 


device for everyday use in the measurement 
of high voltage by deflecting a stream of 
free-moving positively charged particles 
has not been developed, the nuclear phys- 
icist has used such a device in his work for 
a number of years. To help him on his way 
in measuring high voltages he has estab- 

lished the High Voltage Scale. A short di- 
gression in explanation of how this was 
accomplished seems in order prior to a 
presentation of the contribution of the 
free-moving charged-particle deflection- 
method in this work. 

The study of atom physical phenomena 

Fig. 17. — Brooks attracted disk electrometer. 



led to the physicists' need for a voltage 
reference standard of the order of 1,000,000 
times the voltage of the standard cell. 

Historically, one may go back to Planck's 
equation hv = Ve wherein h is Planck's 
constant, v is the maximum frequency cor- 
responding to the maximum accelerating 
voltage V, and e is the value of the ele- 
mentary charge. Duane and Hunt (60) 
used a steady high-voltage storage battery 
in order to evaluate h from a careful meas- 
urement of the accelerating voltage V and 
the corresponding maximum frequency 
limit of the continuous X-ray spectrum 
produced. Since those measurements were 
made the value of h has been well estab- 
lished by other methods. Thus this relation 
affords a method of determining V by a 
measurement of the maximum frequency of 
emitted continuous X-radiation. Little use 
has been made of this method as a high- 
voltage reference standard largely because 
of the accuracy required in evaluating the 
frequency. With the advent of work on 
nuclear disintegration a pressing need for 
high voltage reference standards became 
apparent as the usual extrapolation meth- 
ods of measuring V, the particle accelerat- 
ing voltage, were both cumbersome and 
none too reliable. In early bombardment 
work it was found that resonance radiation 
of gamma rays occurred within a rather 
narrow range of the voltage used in accel- 
erating the protons. The early careful 
measurements by Tuve, Hafstad, and Hey- 
denburg (11) of the voltage at which these 
radiations occurred provided the basis for 
their adoption of certain values of voltage 
corresponding to the resonance gamma ra- 
diation for selected elements as fixed points 
on the High Voltage Scale. 

The measurement process for establishing 
values for the fixed points on the High Volt- 
age Scale consists in: 

(1) Providing a source of protons which are ac- 
celerated by a carefully measured adjustable volt- 
age V. 

(2) Directing this beam of protons on a target 
of a selected element such as lithium or one of its 

(3) Measuring the gamma-ray resonance radia- 
tion effects by means of Geiger-Mueller counters 
or similar devices which permit a quantitative 
measurement of this radiation as a function of the 
accelerating voltage. 

From plotted curves of gamma radiation 
against accelerating voltage (proton en- 
ergy) the rather sharp maximum of gamma 
ray effect fixes the value of voltage chosen 
as the resonance voltage. By selecting 
elements for these reactions in which a 
single sharp maximum occurs and by care- 
fully determining the corresponding volt- 
ages in terms of the standard cell, the volt- 
ages at which particular reactions occur 
were established as reference points on the 
High Voltage Scale. In a similar manner 
but using a boron fluoride ionization cham- 
ber (paraffin surrounding the target tube 
and nearby chamber) instead of the Geiger- 
Mueller counter, neutron counts may be 
made for similar reactions in which neu- 
trons are ejected from the bombarded tar- 
get by voltage accelerated protons. In this 
case points for the scale are determined by 
gradually reducing the accelerating voltage 
V and choosing that voltage at which there 
is an abrupt decrease in ionization chamber 
current as the limiting voltage for ejection 
of neutrons is reached. 

Early experimental values of voltage cor- 
responding to the resonant radiation in the 
bombardment of lithium by protons was 
fixed at 440 kv for Li (py) with a probable 
error of 2 percent and a relative accuracy 
of 1 percent by Tuve, Hafstad, and Hey- 
denburg (11) of the Carnegie Department 
of Terrestrial Magnetism. A redetermina- 
tion of voltage values for some of the fixed 
points was reported in 1944 by Hanson and 
Benedict (59), of the University of Wiscon- 
sin. Their electrostatic analyzer (a device 
for determining the value of the electro- 
static field at right angles to a beam of 
charged particles that will deflect the beam 
a given amount) was carefully constructed 
so that its deflection constant could be 
computed from dimensions as a check on 
values determined experimentally. Their 
experimental method employed an elec- 
tron beam in place of the proton beam 
used later in their evaluation of the fixed 
points. The use of a low-voltage electron 
beam (accelerating voltages of 8 to 20 
kv) and of small deflecting voltages (150 
to 360 volts) permitted precise voltage 
measurements and higher precision for de- 
termining the deflection constant of their 

Feb. 15, 1948 defandorf: the measurement of high voltage 


analyzer than the absolute method consist- 
ing of a computation of the deflection con- 
stant using carefully measured dimensions. 
Various refinements including the use of a 
highly stabilized source for the deflecting 
voltage and automatic regulation of the 
ion-accelerating voltage reduced fluctua- 
tions and contributed to highly precise 
measurements of the fixed points on the 
High Voltage Scale. Hanson and Benedict 
consider the following values expressed in 
terms of the Million Electron Volt scale 
to be accurate to 0.3 percent: IA(py) 
0.4465, F(py) 0.877, Li(pn) 1.883, Be{pn) 
2.058. Their apparatus gave relative values 
agreeing to better than 0.1 percent. Their 
values are seen to agree within the toler- 
ances given for the values of Tuve, Haf- 
stad, and Heydenburg although they are 
approximately 1.5 percent higher. 

The establishment of the High Voltage 
Scale based on a phenomenon unaffected by 
temperature, pressure, and humidity except 
as their abnormalities plague the collateral 
work of the investigator, thus represents a 
distinct step forward in the process of better 
measurements and standards for high volt- 



The principal methods in use for measur- 
ing high voltages have been outlined 
through the discussion of a number of 
devices. Mention was first made of the 
correlation of different methods of high- 
voltage measurements through extrapola- 
tion techniques in which the standard cell 
was used as the primary standard of volt- 
age. Absolute high-voltage electrometers 
were later discussed and the good agree- 
ment between their independently deter- 
mined values when compared with a stand- 
ard-cell voltage-extrapolation method was 
noted. Concluding remarks cited the use of 
both the standard-cell voltage extrapola- 
tion-technique and a less precise absolute 
method in establishing the voltage values 
for certain fixed points on the High Voltage 

At the moment it appears that high 
voltages as we know them in the laboratory 
stop with values of the order of 10,000,000 
volts. On the other hand, charged particles 

come to us from space or may be accelerated 
in the laboratory by resonance techniques 
to have electron velocity equivalents ap- 
proaching one hundred times that value. 
Presumably the future holds in store some 
new insulating arrangement for the col- 
lector of the high energy particles that man 
is able to produce — an arrangement that 
will permit the collector to build up to the 
unbelievable potentials we think of when 
we hear of a new machine in the rumor or 
blueprint stage that will produce 1,000 
million electron-volt particles. Just what 
such voltages will be used for and how they 
will be measured I wish to leave as part of 
your field of conjecture. 


(1) Rayner. Journ. Inst. Electr. Eng. 59: 

138. 1921. 

(2) A. Rosa. Bull. Bur. Stand. 3: 43. 

B. Harris. Bur. Stand. Journ. Res. 3: 
445. 1929. 

(3) Keinath. Die Technik elektrischer Mess- 

gerate2: 17. 1928. 

(4) Chubb. Proc. Inst. Electr. Eng. 35: 

121. 1916. 

(5) Palm. Elektrotechn. Zeitschr. 47: 873, 

904. 1926. 

(6) Konig. Helv. Phys. Acta 1929: 357. 

(7) Haefely et Cie, Basle. German Patent 

No. 394014. 1923. 

(8) Van Cauwenberghe and Landsberg. 

Bull. Techn. Assoc. Ing. Bruxelles 4. 

(9) Davis, Bowlder, and Standring. 

Journ. Inst. Electr. Eng. 68: 1222. 

(10) Taylor. Bur. Stand. Journ. Res. 5: 

609. 1930. 

(11) A. Tuve, Hafstad, and Heydenburg. 

Phys. Rev. 50: 504. 1936. 

B. Hafstad, Heydenburg, and Tuve. 

Phys. Rev. 53: 230. 1938. 

C. Heydenburg, Hafstad, and Tuye. 

Phys. Rev. 56: 1078. 1939. 

(12) A. Silsbee. Bur. Stand. Sci. Pap. 20: 

489. 1925. 

B. Davis. Journ. Inst. Electr. Eng. 

69: 1028. 1931. 

C. de la Gorce. Rev. Gen. d'Electr. 

29: 427. 1931. 

(13) Weller. Trans. Amer. Inst. Electr. 

Eng. 48: 790. 1929. 

(14) A. Elsner. Arch, fur Elektr. 36: 329. 

B. Hohl. Arch, fur Elektr. 35: 663. 

(15) Palm. Zeitschr. fur techn. Phys. 14: 

. 390. 1933. 



(16) Bousman and Ten Bboeck. Trans. 

Amer. Inst. Electr. Eng. 62: 541. 

(17) Churcher and Dannatt. Journ. Inst. 

Electr. Eng. 69: 1019. 1931. 

(18) Bellaschi. Trans. Amer. Inst. Electr. 

Eng. 52: 544. 1933. 

(19) Referred to in Bur. Stand. Journ. Res. 1: 

610. 192S. 

(20) Silsbee and Defandorf. Bur. Stand. 

Journ. Res. 20: 317. 1938. 

(21) Pfiffner.' Elektrotechn. Zeitschr. 47: 

44. 1926. 

(22) A. Keinath. Siemens Zeitschr. 8: 629. 


B. Wirz. Arch, fur Elektr. 21: 563. 


C. Camilli. Gen. Electr. Rev. 39: 95. 


(23) Imhof. Schweiz. Electr. Ver. Bull. 19: 

741. 1928. 

(24) KiicHLER. Elektrotechn. Zeitschr. 58: 

203. 1937. 

(25) A. Wellings and Mortloch. Journ. 

Inst. Electr. Eng. 79: 577. 1936. 
B. Probst. Elektrotechn. Zeitschr. 28: 
750. 1924. 

(26) Hendricks, Hubbard, and Vallin. 

Gen. Electr. Rev. 42: 420. 1939. 

(27) A. Cahall. U. S. Patent No. 1979096. 

B. Hendricks. Gen. Electr. Rev. 46: 
477. 1943. 

(28) Kirkpatrick. Electr. Eng. 51: 863. 


(29) Workman and Holzer. Rev. Sci. Instr. 

10: 160. 1939. 

(30) Brooks, Defandorf, and Silsbee. 

Bur. Stand. Journ. Res. 20: 253. 1938. 

(31) Herb, Parkinson, and Kerst. Rev. 

Sci. Instr. 6: 261. 1935. Phys. Rev. 
51: 75. 1937. 

(32) Kirkpatrick. Rev. Sci. Instr. 3: 1, 

430. 1932; 5: 33. 1934. 

(33) ASA American Standards for Measure- 

ment of Test Voltage in Dielectric 
Tests, C68.1-1942. 

(34) VerPlanck. Amer. Inst. Electr. Eng. 

Techn. Pap. 41-4. 

(35) Loeb. Fundamental processes of electrical 

discharge in gases. New York, 1939. 

(36) Meek. Journ. Franklin Inst. 230: 229. 


(37) Meek. Journ. Inst. Electr. Eng. 93 

(11-32): 97. 1946. 

(38) VerPlanck. Trans. Amer. Inst. Electr. 

Eng. 57:45. 1938. 













Bellaschi and Teague. Electr. Journ. 

32:56. 1935. 
Thornton, Waters, and Thompson. 

Journ. Inst. Electr. Eng. 69: 533. 

Abraham and Villard. Journ. de Phys. 

1911: 525. 
Schumann and Klemm. Arch, fur Elektr. 

12:553. 1923. 
Starke and Schroeder. Arch, fur 

Elektr. 20: 115. 1928. 
Imhof. Arch, fur Elektr. 23 : 258. 1929. 
Th. Wulf. Phys. Zeitschr. 31: 315. 

Hueter. Elektrotechn. Zeitschr. 55: 

833. 1934; 56: 1319. 1935. 
Sorensen, Hobson, and Ramo. Electr. 

Eng. 54: 651. 1935. 
Sorensen and Ramo. Electr. Eng. 55: 

444. 1936. 

Rogowski. Arch, fur Elektr. 25: 521. 

Rogowski and Bocker. Arch, fur 

Elektr. 32: 44. 1938. Elektrotechn. 

Zeitschr. 59: 123. 1938. 
Bocker. Arch, fur Elektr. 33: 801. 

1939. Elektrotechn. Zeitschr. 61 : 729. 

Kelvin. Papers on electrostatics and 

electr omagnelism, ed. 2, sect. 360: 287. 

London, 1884. 

A. Thornton and Thompson. Journ. 

Inst. Electr. Eng. 71: 1. 1932. 

B. Bruce. Journ. Inst. Electr. Eng. 94 

(2): 129. 1947. (Published sub- 
sequent to delivery of this paper.) 
Snow. Bur. Stand. Journ. Res. 1: 513. 

Nacken. Arch, fur Elektr. 33: 60. 

1939; 36: 678. 1942. 
Tschernyscheff. Phvs. Zeitschr. 11: 

445. 1910. 

Palm. Zeitschr. fiir techn. Physik 1: 
137. 1920. 

A. Charlton and Cooper. Gen. Electr. 

Rev. 40:438. 1937. 

B. Trump, Safford, and Cloud. 

Trans. Amer. Inst. Electr. Eng. 60: 
132. 1941. 

C. Schumb. Phys. Rev. 69: 692. 1946. 
Hanson and Benedict. Phvs. Rev. 65 : 

33. 1944. 
Duane and Hunt. Phvs. Rev. 6: 166. 

Winkelbrandt. Arch, fur Elektr. 31: 

672. 1937. 

Feb. 15, 1948 bates and smith: standardization of the pn scale 


CHEMISTRY. — Standardization of the pH scale. 1 Roger G. Bates and Edgar 
Reynolds Smith, National Bureau of Standards. 

The pH is not a definite thermodynamic 
quantity, and the lack of a universally 
accepted definition of the pH scale has led 
to an unfortunate state of confusion. The 
two scales of pH in common use at the pres- 
ent time differ by 0.04 unit. Some workers 
are using one definition for computations 
and measuring a different quantity. To 
correct this situation it is of primary impor- 
tance that a single scale be generally adopted 


Commercial pH meters of the glass-elec- 
trode type are now found in nearly every 
laboratory where chemical analyses and 
tests are performed. These instruments 
must be calibrated from time to time with 
buffer solutions of assigned pH. Two difficul- 
ties are encountered in any attempt to 
measure pH accurately. First, neither the 
concentration nor the activity of hydrogen 
ion in a buffer solution can be exactly and 
uniquely determined. In the second place, 
the potentials at the liquid junctions be- 
tween the various standard and unknown 
solutions and the bridge solution of the 
reference electrode can never be perfectly 
matched or eliminated. Measured pH 
values will still be uncertain for this reason, 
even when exact standards are available. 
Nevertheless, the larger differences due to 
lack of agreement on a fundamental defini- 
tion and on a method of calculation can be 
avoided by the co-operation of all who 
determine pH. As a means to this end, the 
National Bureau of Standards proposes the 
adoption of a standard scale of pH having 
reference points based upon certain repro- 
ducible buffers. In this discussion, the ad- 
vantages and limitations of the several 
common pH scales will be considered, and 
the nature of the problems involved in an 
evaluation of the hydrogen-ion activity of 
the standard buffer solutions from electro- 
motive-force measurements will be indi- 

S. P. L. S0rensen first proposed the use of 
the hydrogen-ion exponent (Potenz) to 
facilitate the designation of the extremely 

1 Received September 2, 1947. 

small concentrations of hydrogen ion of 
physiological significance (#). This unit of 
acidity will be called pcH. Its definition is 

pcH=-log c H ;pcH=-log m H , (1) 

where c represents volume concentration 
(molarity or normality) and m is molality. 
The hydrogen-ion concentration, it was 
supposed, could be obtained experimentally 
by measurement of the electromotive force 
(emf, E) of a suitable galvanic cell. S0rensen 
chose the cell, 

Pt; H 2 (g), Solution 




Hg 2 Cl 2 (s);Hg, 


for the practical determination of pcH. The 
vertical lines represent liquid junctions. Ion 
transfer across these boundaries gives rise to 
an unknown liquid-junction potential, Ej, 
which was partially eliminated by the ex- 
trapolation procedure suggested by Bjer- 
rum (3). 

With the aid of standard solutions com- 
posed of hydrochloric acid with and without 
sodium or potassium chloride, S0rensen 
determined E°, the "standard potential" or 
"normal potential" of cell (2). The hydro- 
gen-ion concentration of these standard 
solutions was taken to be the product of the 
concentration of hydrochloric acid and the 
classical degree of dissociation derived from 
measurements of electrolytic conductance. 
By definition, E° is the potential of cell (2) 
when the hydrogen-ion concentration of the 
solution is 1 normal. S0rensen's value of E° 
is 0.3380 at 18° C (2, 4, 5, 6). At 25° C, the 
standard potential is 0.3376 (6). It was in- 
tended that the pH at 25° C of "unknown" 
solutions should be computed by the simple 


p S H= , (3) 


where E is corrected insofar as possible for 
the undesired potential at the liquid junc- 
tion between the solution and 3.5-ilf potas- 



sium chloride. This experimental pll value 
defined by S0rensen's value of E° is not, as 
he intended, pell. It is designated psH in 
equation (3). The constant 0.0591 in the 
denominator is the value of 2.303 RT/F at 
25° C, where R, T, and F represent the gas 
constant, absolute temperature, and fara- 
day, respectively. 

In the light of modern concepts, it is 
clear that the laws of ideal solutions can not 
justifiably be employed in this way. Nor 
will the classical degree of dissociation 
yield, as S0rensen assumed, the hydrogen- 
ion concentration of mixtures of strong 
electrolytes. Activity now replaces concen- 
tration in the equations for the emf . Indeed, 
at a temperature of 25° C, 

log a H = 



where a represents activity. The standard 
potential, E 0> ', is referred to unit activity 
instead of normality and includes the 
constant chloride-ion activity in the vicin- 
ity of the calomel electrode. The activity, a, 
is the product of concentration, c or m, and 
an activity coefficient, /. If Ej could be 
evaluated, the exact hydrogen-ion activity 
would be forthcoming. Unfortunately, the 







2° 4.65 





pafl — - 












0.06 0.( 


Fig. 1. — pH of acetate buffer solutions at 25° C 
on four different scales as a function of concen- 

liquid-junction potential can not be calcu- 
lated without a knowledge not only of /h 
and other activity coefficients but also of 
the compositions of the multitude of transi- 
tion layers of which the boundary is com- 
posed. For this reason, the activity of a 
single ionic species is without physical 
reality (7). 

The new concept of the thermodynamics 
of cell (2) led to the definition of paR by 
S0rensen and Linderstr0m-Lang (5): 

paR = —log a H . 


Although paR can not be determined by 
thermodynamic methods, its character can 
be simply and unequivocally defined in 
terms of measurable mean activity coeffi- 
cients. The paR scale has likewise been 
defined in terms of the thermodynamic 
ionization constants of weak acids in dilute 
buffer solutions (8,9, 10, 11). 

The pR values, on these three scales, of 
buffer solutions composed of equal molar- 
ities of acetic acid and sodium acetate are 
compared in Fig. 1. At infinite dilution, 
paK and pcR are equal, for the activity 
coefficient becomes unity by definition in 
that limit. However, the S0rensen value, 
psR, remains lower by about 0.04 unit than 
paR at all concentrations. The pwR, defined 
as —log (a H /ci), changes but little with dilu- 
tion. This unit will receive further mention. 
When it is desired to convert psR to paR, 
the following relationship will serve with an 
uncertainty probably less than 0.02 unit: 

paR = psR+0M. 


The S0rensen scale, or psR, is probably 
the scale most widely used today. The pR 
values given in the well-known monograph 
of Clark (6) correspond to this scale. Yet 
psR is neither — log a H nor — log c H . It usually 
lies, in fact, between these. When a pR 
meter is standardized on this scale, all 
measured values must be corrected to yield 
a quantity that can be employed in equilib- 
rium computations, if they are to have 
quantitative significance. The hydrogen- 
ion concentration possesses the physical 
reality that the activity lacks. A knowledge 
of the hydrogen-ion concentration would be 
useful, but this quantity can not readily be 
determined by an emf measurement with 

Feb. 15, 1948 bates and smith: standardization of the pn scale 


the pH meter. For this reason, the activity 
scale has been chosen as the most practical 
for general use. We shall now consider the 
assignment of paR values to the buffer 
solutions that will serve as fixed reference 
points on this scale. 

Cells without liquid junction are not only 
somewhat more advantageous from a theo- 
retical standpoint than are cells with liquid 
junction, but they are also more reproduci- 
ble and can readily be measured at different 
temperatures. They are, however, usually 
impractical for routine measurements. 
Efforts have been made at the National 
Bureau of Standards to use emf measure- 
ments of cells without liquid junction to 
establish standards of hydrogen-ion activity 
with which to calibrate the pH meter. Each 
emf measurement of the hydrogen-silver 
chloride cell, 

monobasic acid or an acid anion, and pYL 
for the dissociation equilibrium is known, 
f r is also obtained unequivocally: 

log f r = log 




= pK — log 

m H A 

m A 

flog (/h/ciWh). (9) 

Pt; H 2 (g, 1 atm), Buffer solution, 


The quantity —log (/h/ci^h) will be 
termed pwH. Inasmuch as pwH is thermo- 
dynamically exact, its use as a unit of acid- 
ity has been suggested {12, 13). The pwB. of 
a buffer solution composed of a weak mono- 
basic acid, such as acetic acid, and its salt, 
however, changes but slightly with rela- 
tively large variations in hydrogen-ion con- 
centration, as shown in Fig. 1. For this 
reason, — log (/h/ci^h) is often not a useful 
unit of acidity in spite of the fact that it 
retains its significance at all ionic strengths. 
The paH is derived from pwH by adding the 
logarithm of an ionic activity coefficient: 

yields a value of —log (/h/ci^h), if the 
standard potential, E°, and the molality, 
m, of chloride ion in the solution are known, 
by the following equation: 

pdR = pwK+\ogf C \ 

= pK — \og 

m A 

log/ r +log/ci. (10) 

-log C/h/ciWh) 


flog mci. (8) 

If the buffer solution is composed of a weak 
acid, HA, and its salt, where HA is either a 

Thermodynamics can offer no help in 
estimating the activity coefficient of chloride 
ion in equation (10). For this reason it is 
usually necessary to resort to theoretical 
equations or to assumed relationships 

Table 1 — poH of Buffer Solutions without Chloride at 25° C. Computed from the EMF of Cell (7) 

Buffer system* 


of each 



piH for — 

Pj h 

p 5 H 

p t B. 

p 5 H 

paH, cells 

with liquid 


at =4 

at =6 

Acetic acid + 

/ 0.01 
\ 0.1 


f 0.01 
i 0.02 
| 0.05 
1 0.1 

' 0.005 


< 0.025 


f 0.01 
\ 0.025 































4.700; 4.714 
4.640; 4.645 

Acid potassium phthalate . 

Acid potassium 
phthalate +dipotassium 

4.000; 4.010 

Potassium dihydrogen 
hydrogen phosphate 




* When two components are present, their concentrations are equal. 



among the ionic activity coefficients in com- 
binations that can be experimentally deter- 
mined, such as mean activity coefficients or 
f r of equation (9). These methods give 
equivalent results in very dilute solutions. 
Some of these assumptions will now be con- 

The equations of Debye and Huckel (14, 
15) represent observed values of the mean 
activity coefficients of strong electrolytes 
with considerable success. Thus the Huckel 

AziWn _ 


might be employed to evaluate the activity 
coefficient of an ion i of valence z { . In this 
equation, A and B are constants at a par- 
ticular temperature for the water medium, 
whereas a z - and £»• are parameters character- 
istic of the mixture of ions. At ionic 
strengths, /x, below 0.1, the linear term can 
often be compensated by a small change in 
di and the logarithm of the activity coeffi- 
cient expressed satisfactorily by the first 
term on the right alone. The magnitude of 
a. is of the same order as the ionic diameter 
in angstroms. Although the numerical 
value of this parameter lies between 3.5 and 
6.5 for many strong electrolytes, there is no 
known basis for selecting the correct value 
of a» for a single ionic species. 

Inasmuch as cell (7) is reversible to hy- 
drogen and chloride ions, it would not be 
unreasonable to assume that the activity 
coefficient of chloride ion is about equal to 
the mean activity coefficient of hydrochloric 
acid in the buffer solution or in a mixture of 
strong electrolytes of the same ionic 
strength and composition with respect to 
cations. Unfortunately, these mean activity 
coefficients are often unknown. The mean 
activity coefficient of hydrochloric acid in 
its pure aqueous solution is well established, 
however, over a wide range of concen- 
trations (16, 17, 18). A practical scale of 
paH can be defined by assuming the equal- 
ity of /ci in the buffer mixture and / H ci in a 
solution of hydrochloric acid of the same 
ionic strength. 

Another possible approach is the sepa- 
ration of the measured f r into /ci and /ha//a- 

Practical considerations limit this compu- 
tation to buffer systems in which HA is an 
anion, that is, to systems where A bears a 
charge different from CI. The partition of 
f r might reasonably be based upon valence 
relationships valid in very dilute solutions, 

/ci=/a-=/a— 1/4 =/a- 



where A~, A , and A represent anions 
with 1, 2, and 3 negative charges. 

Two other separation formulas are most 
readily described in terms of the parameters 
of equation (11). When the acid, HA, is a 
singly charged anion, it has been found that 
f r °, the limit of f r in chloride-free solutions, 
can be expressed by 

l0g/r° = 





One method of obtaining / C i assumes that a 
and j8° can be identified with a r - and & for 
the computation of /ci by equation (11). 
The other method identifies a° with a; of 
equation (11) and drops the linear or "salt- 
effect" term in computing f C \ in a buffer 
solution that contains no chloride. 

These five assumptions lead to five differ- 
ent pall scales (19). When -log f C \ is com- 
puted by the first term on the right of equa- 
tion (11) with a value of a { not derived from 
experimental data, the scale will be called 
PiH. The p 2 H is that unit obtained by set- 
ting /ci equal to / H ci in a solution of the 
same ionic strength as the buffer solution. 
Partition of f r according to the relationships 
assumed in equation (12) yields p 3 H. Iden- 
tification of a and /3° with a { and & leads to 
the unit designated p 4 H, whereas omission 
of the salt-effect term gives p 5 H. The paK 
of buffer solutions without added chloride is 
thus expressed in terms of experimentally 
defined quantities by the following equations : 

p 1 H^(^H)°-.4 v / M/(l+5aVM). (14) 

p 2 H = (pt0H)°+log/ H ci. (15) 

p 3 H = (pwH) -l/21og/ r <>. (16) 

pJl^(pwHy-l/2 log//>+3/2 £V (17) 

Vb R = (pwH) Q - 1/2 log / r °+ 1/2 /3V. (18) 

The relationship among these last three 

Feb. 15, 1948 bates and smith: standardization of the pn scale 


paH scales is evidently given by 

p 4 H = p 6 H+ j SV = P3H+3/2 j 9V. (19) 

In each of these equations the superscript 
zero indicates that the effect of chloride has 
been removed by extrapolation to a pure 
buffer solution without chloride. 

The paK at 25° C of several buffer solu- 
tions on these five scales is listed in Table 1 
and compared, where possible, with the paK 
derived from measurements of cells with 
liquid junction reported by Hitchcock and 
Taylor (9) and by Maclnnes, Belcher, and 
Shedlovsky (11). Inasmuch as the choice of 
a { in equation (14) is partially arbitrary, 
two piH. values, calculated with 4 and 6 for 
a { , are given. When the ionic strength is 
0.01, these differ by only 0.003 unit. The 
pwH of the acetate, phthalate, phosphate, 
and borax solutions was derived from pub- 
lished emf data (19, 20, 21, 22). 

It must be emphasized that thermody- 
namics offers as much, or as little, support 
for the choice of one paH scale as another. 
One cannot state categorically that a partic- 
ular method of computation is wrong and 
another right. The assumptions can only be 
compared with respect to their reasonable- 
ness. Chloride ion evidently plays a unique 
role in these pH equations. None of these 
formulas can qualify as adequate unless it 
furnishes the same paH. for a given buffer 
solution in the absence of sodium bromide 
or sodium iodide, for example, as it gives in 
the limit of zero concentration of sodium 
chloride. Electromotive-force measurements 
of hydrogen-silver halide cells containing 
phosphate buffer solutions with added 
sodium chloride, sodium bromide, and so- 
dium iodide (23) offer an interesting, though 
not exhaustive, test of the adequacy of 
these five methods of computation (19). 
Both pwYL and log f r were obtained for 
varying ratios of halide to phosphate. The 
limiting values in the phosphate buffer so- 
lutions without halide were found by extra- 
polation. By fitting/,. to equation (13), a 
and jS° were determined as well. 

The pdR of equimolal phosphate buffer 
solutions on the p 2 H, p 3 H, p 4 H, and p 5 H 
scales was computed by equations (15), 
(16), (17), and (18). The values on the last 
four scales are shown in Fig. 2 as a function 

of ionic strength. The lines representing 
these scales are labeled 2, 3, 4, and 5. The 
results derived from bromide cells are 
marked with a single prime and those from 
iodide cells with a double prime. Lines un- 
marked save for the identif jang figure repre- 
sent data from the chloride cells. The dots 
(curve 5) are the values of the NBS scale. 
The upper and lower dashed lines locate the 
PiH curve when a z - is arbitrarily assigned the 
extreme values of 8 and 3 for the computa- 
tion. The molality of each phosphate salt 
is one quarter of the ionic strength. 

The three sets of P2H values computed by 
equations of the form of equation (15) from 
— log (/h/xWh), where X represents halide, 
and the activity coefficient of the correspond- 
ing halogen acids agree well among them- 
selves but are from 0.01 to 0.03 unit higher 
than curve 5 at ionic strengths between 0.1 
and 0.2. It is noteworthy, however, that all 
methods of calculation give essentially the 
same paH at low ionic strengths. The course 
of the true — log a H curve can never be ascer- 
tained. Nevertheless, at ionic strengths of 


0.08 0.12 


Fig. 2. — p 2 H, P3H, p 4 H, and p 3 H of phosphate 
buffer solutions as a function of ionic strength. 
Curves representing the four scales are labeled 
2, 3, 4, and 5. The values were derived from the 
emf of cells with silver-silver chloride electrodes 
(unprimed), silver-silver bromide electrodes (sin- 
gle prime), and silver-silver iodide electrodes 
(double prime). The dashed lines indicate the 
course of the curve of piH for a\ values of 8 
(upper line) and 3 (lower line). The dots (curve 
5) are the NBS values. 



0.01 or below, where all of these reasonable 
assumptions give practically identical re- 
sults, the paH can be said to possess some 
measure of thermodynamic significance. 

It is evident that the paK is markedly 
affected at high or even moderate ionic 
strengths by the assumption used in its 
evaluation. For this reason, primary stand- 
ards of hydrogen-ion activity should be 
solutions of low salt concentration. Al- 
though the true activity is unknown at the 
higher ionic strengths, the NBS pH stand- 
ards are consistent with one another over 
the pH. range 4 to 9.2 at all concentrations 
to which the pR has been assigned. The 
NBS scale is a true scale of activity at low 
concentrations. At the higher ionic strengths 
it is best regarded as a self -consistent scale 
which, though based upon activity, per- 
force embraces an assumption not subject 
to experimental proof. This scale of pH ap- 
pears to be the most convenient and useful 
to adopt as standard. 


(1) Smith, E. R., and Bates, R. G. Comptes 

rendus de la reprise de contact du bureau, 
du conseil, et des commissions, a Vissue 
de la seconde guerre mondiale, p. 63. 
International Union of Chemistry, 
London, July 1946. 

(2) S0rensen, S. P. L. Compt. Rend. Lab. 

Carlsberg 8: 1. 1909. 

(3) Bjerrum, N. Zeitschr. Physik. Chem. 

53: 428. 1905. 

(4) S^rensen, S. P. L. Ergeb. Physiol. 12: 

393. 1912. 

(5) S^rensen, S. P. L., and Linderstr^m- 

Lang, K. Compt. Rend. Trav. Lab. 
Carlsberg 15, No. 6. 1924. 

(6) Clark, W. M. The determination of 

hydrogen ions, ed. 3. Baltimore, 1928. 

(7) Taylor, P. B. Journ. Phys. Chem. 31: 

1478. 1927. 

(8) Cohn, E. J., Heyroth, F. F., and Men- 

kin, M. F. Journ. Amer. Chem. Soc. 
50: 696. 1928. 

(9) Hitchcock, D. I., and Taylor, A. C. 

Journ. Amer. Chem. Soc. 59: 1812. 
1937; 60: 2710. 1938. 

(10) Kauko, Y., and Airola, A. Zeitschr. 

Physik. Chem. A178: 437. 1937. 

(11) MacInnes, D. A., Belcher, D., and 

Shedlovsky, T. Journ. Amer. Chem. 
Soc. 60: 1094. 1938. 

(12) Guggenheim, E. A. Journ. Phys. Chem. 

34: 1758. 1930. 

(13) Hitchcock, D. I. Journ. Amer. Chem. 

Soc. 58: 855. 1936; 59: 2753. 1937. 

(14) Debye, P., and Huckel, E. Physik. 

Zeitschr. 24: 185. 1923. 

(15) Huckel, E. Physik. Zeitschr. 26: 93. 


(16) Harned, H. S., and Ehlers, R. W. 

Journ. Amer. Chem. Soc. 55: 2179. 

(17) Shedlovsky, T., and MacInnes, D. A. 

Journ. Amer. Chem. Soc. 58: 1970. 

(18) Akerlof, G., and Teare, J. W. Journ. 

Amer. Chem. Soc. 59: 1855. 1937. 

(19) Bates, R. G. Chem. Rev., in press. 

(20) Hamer, W. J., and Acree, S. F. Journ. 

Res. Nat. Bur. Standards 32: 215. 
1944; 35: 381. 1945. 

(21) Bates, R. G., and Acree, S. F. Journ. 

Res. Nat. Bur. Standards 34: 373. 1945. 

(22) Manov, G. G., DeLollis, N. J., Lind- 

vall, P. W., and Acree, S. F. Journ. 
Res. Nat. Bur. Standards 36: 543. 1946. 

(23) Bates, R. G. Journ. Res. Nat. Bur. 

Standards 39: 411. 1947. 

ANTHROPOLOGY. — The true form of the cranial deformity originally described 
under the name "tete trilobee. ,n T. D. Stewart, U. S. National Museum. 

Among the earliest accounts of deformed 
skulls from America is one by Louis- Andre 
Gosse, a Swiss physician-anthropologist, 
published in 1855. Included in this essay, 
which is concerned primarily with the classi- 
fication of deformity types, is the first de- 
scription of a skull from Isla de los Sacri- 
ficios, a tiny island 5.5 km southeast of the 
port of Veracruz, Mexico. Gosse's descrip- 
tion of this skull is of historical importance 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received September 
9, 1947. 

because the artificial distortion which it 
exhibited which he characterized as "tete 
trilobee" has influenced all subsequent gen- 
eralizations regarding the distribution of de- 
formity types in Middle America. Thus in 
Dingwall's book on cranial deformity 
(1931, p. 154) it is stated that— 

The skulls from Isla de los Sacrificios, near 
Veracruz, where the Spaniards first saw the gory 
remains of human sacrifice are unusually dis- 
torted, and Salas [1921, quoting Gosse] describes 
them as trilobed on account of the depressions 
doubtless left by the constricting bandages. This 
again suggests that at least two important forms 

Feb. 15, 1948 stewart: the true form of the "tete trilobee' 


of cranial deformation were known in Central 
America from early times, that produced by 
boards and that produced by bandages. 

Also, Imbelloni (1934, p. 65), in plotting out 
the regions of deformity throughout the 
hemisphere, says that — 

[The Quiche-Huasteca] region presents deformi- 
ties of the type [tabular erecto], with the peculiarity 
of transverse grooves that, according to Gosse, 
facilitated the carrying of burdens (they are what 
Gosse calls trilobed). It comprises the coast of 
Mexico corresponding to the province of Vera- 
cruz (with the nearby Isla de los Sacrificios) and 
Guatemala. (Free translation.) 

Recently (1943) a careful archeological 
investigation was made on Isla de los Sacri- 
ficios by the Mexican archeologist Wilfrido 
du Solier in the course of which several 
skulls were encountered. Through the 
kindness of du Solier I was able to examine 
these skulls during my visit to Mexico City 
in September 1946. So different from the 
classical description of Gosse is the type of 
deformity in these new specimens that I 
have reviewed the whole subject with the 
interesting results that follow. 


According to Gosse, his description of the 
Sacrificios type of deformity was based upon 
the cast of one of several male skulls brought 
to France and deposited in the " Anthro- 
pological Museum of Paris" by a French 
naval officer named "Reymond." The cast 
was made by a Paris physician, Guy the 
Elder, and was seen by Gosse in the museum 
in Geneva. The peculiar trilobed shape ex- 
hibited by the cast is clearly shown in 
Gosse's drawings (Fig. 1). 

In 1861, at a meeting of the Anthropologi- 
cal Society of Paris, Gosse again referred 
to this type of deformity; this time in con- 
nection with his paper on a deformed skull 
from Ghovel, Mexico. Here he uses the 
term "trilobee" synonymously with "occi- 
pito-sincipito-frontale." In explanation he 
says (1861, pp. 576-577): 

The impressions left on the skull seem to indi- 
cate that the trilobed form was obtained by means 
of a thick compress, narrow and long, extending 
from the neck to the sinciput, that produced a 
deep depression along the median line of the 
occipital squama and along the posterior half of 
the sagittal suture, thus dividing the back of 

the head into two lobes. Moreover, one or two 
small compresses were applied on the frontal and 
the whole was held in place by the aid of two 
bands, one transverse, passing over the sinciput, 
and the other circular, passing around the base 
of the skull. (Free translation.) 

Further information on this type of de- 
formity was supplied by Hamy (1884- 
1891). 2 According to this author, the cast 
seen by Gosse had been considerably altered 
("fortement remanie") and consequently 
the description based thereon should be 
modified in many important points (foot- 
note 2, p. 91; footnote 1, p. 93). Further- 
more, Hamy states that the skull from 
which this cast was made may have come 
from the Dumanoir collection (footnote 2, 
p. 91). This was plausible because Captain 
Dumanoir, who commanded the French 
corvette Ceres, explored Isla de los Sacri- 
ficios in 1841 and discovered sepulchers 
there containing human remains (Mayer, 
1844, p. 96). However, all this is difficult to 
reconcile with Hamy's specific statement 
(p. 91) that the Dumanoir collection had 
not been described before 1884. 

Besides the Dumanoir collection, Hamy 
mentions (p. 91) a collection made by a Dr. 
Fuzier, who excavated at Sacrificios during 
the French occupation (1838) and who gave 
to the museum "a certain number" of more 
or less damaged skulls. Three of these skulls 
are shown in line drawings of the norma 
verticalis in Hamy's plates 10 and 11. No 
other skulls from Sacrificios are specifically 
mentioned in the text or tables of measure- 
ments. A total of 6 specimens, 5 male and 1 
female, without reference to collector, is 
indicated in the table on Hamy's p. 92. 

In the plate illustrating the two skulls 
from the Fuzier collection, which plate I am 
designating no. 10, are featured two other 
skulls. These are stated in the legend to be 
from Sabine Lake and the Reynaud collec- 
tion. Referring to the relevant text (p. 98), 
we find that — 

2 This publication appeared in three parts be- 
tween the dates given. Most bibliographies list 
it as incomplete, perhaps because plate no. 18 is 
missing. It does not appear that this plate ever 
was printed. Moreover, two plates are given the 
number 10, whereas none bears the number 12. I 
have assumed from the wording of the text that 
the plate illustrating the skulls from Sacrificios 
and Sabine should be no. 10. 



Fig. 1.— Two views of the cast of a deformed skull from Isla de los Sacrificios, 
Mexico, as illustrated by Gosse in 1855 (pi. 1, figs. 4 a and 4 b ). 

The skulls offered to the Museum of Natural 
History of Paris by Admiral Reynaud under the 
name of skulls from Sabine Gulf of Mexico do not 
differ sensibly from those of the second type from 
the island of Sacrificios. 3 

Footnotes 3 and 4 on the same page offer 
the additional information that these 
skulls — 

. . . foi a long time have been confused under the 
name of Sacrificios, but their primitive label is 
such that we came to transcribe it. Sabine is the 
name of a gulf and of a river which forms the 
boundary between Texas and Louisiana. 

Our first type of Sacrificios was not encoun- 
tered among these skulls. (Free translation.) 

Line draAvings of one of the skulls said to 
have come from Sabine are shown in Fig. 2. 
Readers will recall that Gosse attributed 
the original of his cast to the collection of a 
naval officer named "Reymond." This 
name is remarkably similar to that of the 
admiral, Reynaud, who, Hamy says, made 
the Sabine collection, which was believed 
for a long time to have come from Sacrifi- 

3 Hamy's second type of deformity corresponds 
to Gosse's "tete trilobee." Hamy's first type is 
based upon a single specimen and differs merely 
in having a greater compression of the frontal 

cios. Could it be, therefore, that one of these 
names is an erroneous spelling and that the 
same collector is referred to? If so, did Gosse 
really describe a skull from Sabine instead 
of from Sacrificios? Or is Hamy in error in 
changing the locality of the collection from 
Sacrificios to Sabine? These are important 
questions, because if it can be proved that 
this deformity type occurs as far north on 
the coast of the Gulf of Mexico as the 
boundary between Texas and Louisiana, 
another cultural link between Mexico and 
the southeastern United States will have 
been established. 

Further search in the literature has 
brought to light two references that clarify 
this subject somewhat: 4 (1) In Larousse's 
Grand dictionnaire universel du XI X e siecle 
(1875, vol. 3, p. 1136) there is a biography of 
a distingusihed French naval officer by the 
name of Aime-Felix-Saint-Elme Reynaud 
who began his maritime career in 1827. In 
1850 he became Captain of a frigate and in 
1864 a Vice- Admiral. This biographical dic- 

4 I am indebted to Mrs. Elizabeth H. Gazin, 
U. S. National Museum librarian, for locating 
these references. 

Feb. 15, 1948 stewart: the true form of the "tete trilobee" 


tionary fails to list a "Reymond." (2) A foot- 
note to a paper by Serres (1855, footnote 1, 
p. 46) reads in part as follows: 

The ideal beauty that the Aztecs sought to 
produce was favored by the normal elongation 
of the bones of the skull and face of this tribe. By 
compression they only exaggerated their normal 
type. This is true also of the skulls which for the 
first time have been given to the museum by Reynaud, 
distinguished officer of the French navy and so far 
found only in the island of the sacrifices, in the Gulf 
of Mexico. All of the bones of the skull and face 
are large, the inverse of the preceding. The com- 
pression thus was exerted in inverse sense; it was 
for the purpose of enlarging the skull and giving it 
the trilobed form that they represent .... (Free 
translation; italics mine.) 

Since Serres' paper was published in the 
same year — 1855 — as that by Gosse, and 
both he and Gosse refer to a collection of 
skulls from Sacrificios given earlier to a 
Paris museum by a French naval officer, it 
seems reasonable to believe that they are 
speaking of the same event. 5 Furthermore, 
since Serres gives this officer's name as 
"Reynaud" and there was a distinguished 

5 Gosse (1855, Paris ed., p. 40) states that Prof. 
Serres gave him permission to examine the speci- 
mens in the anthropological museum. 

French naval officer by that name living at 
that time, whereas Gosse gives this officer's 
name as "Reymond," for which there is no 
corresponding biographical record, it seems 
likely that Gosse misspelled the name. 
Such misspelling is understandable because 
in longhand a's may look like o's, u's like 
n's, and m's like n's. 

What is still not explained is the later 
confusion as to the provenience of this col- 
lection. It seems to have been well estab- 
lished in 1855 that Reynaud obtained it 
from Sacrificios. And yet in 1891 Hanry 
reports the old label as reading "cranes de 
Sabine, gulfe du Mexique." In view of the 
improbability that a French naval officer 
would have an opportunity just prior to 
1855 to carry out archeological excavations 
in the region of Sabine Lake (Texas- 
Louisiana) and in view of the subsequent 
failure to recover other skulls thus deformed 
from either Texas or Louisiana, I believe 
that Ham} r was wrong in attributing the 
Reynaud collection to the Sabine. There- 
fore, I regard these skulls not only as ex- 
hibiting a type of deformity characteristic 
of Sacrificios, as Hamy admits, but as 
actually coming from that place. 

Fig. 2. — Two views of deformed skull no. 5 from the Reynaud collection believed by Hamy to have 
come from Sabine Lake on the Texas-Louisiana boundary. Redrawn and reoriented from Hamy's 
illustrations of 1884-91 (pi. 10, figs. 1 and 5). 




Taking into account this decision, let 
us now summarize what is known about the 
skulls from Sacrificios. All these, except for 
the new finds of du Solier, seem to be in 
Paris and to total at least 15. The early 
collectors, who visited the island around the 
years 1838-1841, were Dumanoir, Fuzier, 
and Reynaud. The only firsthand descrip- 
tions of the deformity are by Gosse (1855) 
and Hamy (1884-1891). These descriptions 
seem to be based upon extreme cases, for 
both Gosse and Hamy admit that the arti- 

ficial characteristics are less pronounced in 
the majority of cases than they have de- 
scribed them. Thus there is good reason to 
believe that the skull shown in Fig. 2, 
which was illustrated originally only be- 
cause it was believed to come from Sabine 
rather than from Sacrificios, is more typi- 
cally deformed. So far as I can discover 
from the descriptions, only one skull has 
the curious projection of the upper part of 
the frontal bone illustrated by Gosse (Fig. 
1). This one specimen seems to have fur- 
nished the excuse for the name "tete tri- 

Fig. 3. — Two views of a deformed skull collected by du Solier in 1943 on Isla de los Sacrificios, 
Mexico. (Photographs supplied by the Museo Nacional of Mexico.) 

Feb. 15, 1948 stewart: the true form of the "tete trilobee' 


lobee" under which this deformity is gen- 
erally known. 

If we ignore this extreme case, the de- 
formity can be recognized by the following 
characteristics: (1) When the skull is held 
in the Frankfort position the vertex or 
highest point is at or near bregma; (2) 
looked at from above, the skull is very 
broad, sometimes broader than long; (3) 
the posterior parietals are pressed down- 
ward and forward, sometimes with ac- 
companying compression of the frontal 
and/or occiput (the latter tending to as- 
sume a vertical plane); and (4) there may 
be more or less of a depression along the 
sagittal suture and behind the coronal su- 
ture. A skull deformed in this manner sug- 
gests that it has had its growth restricted 
along the midline of the vault and conse- 
quently that expansion took place laterally, 
Also, when compression is high upon the 
parietals the skull appears to have a low 

The skulls recovered by du Solier are not 
markedly distorted. Indeed, the lateral 
view of the skull shown in Fig. 3 looks un- 
deformed. This is due to the fact that the 
flattening of the posterior parietals in this 
case is asymmetrical (see vertical view) and 
the least involved side has been presented 
to the camera. As a matter of fact, this 
skull illustrates, if not ideally, the really 
significant element of the Sacrificios de- 
formity type ; namely, the inclined compres- 
sion plane involving the posterior parietals 
and the upper part of the occipital. In ac- 
cordance with current custom I am calling 
this element of the deformity "lambdoid" 
flattening, although it might with more ac- 
curacy be called "obelionic" (Stewart, 
1939). As I have shown elsewhere (Stewart, 
1947; in press), lambdoid flattening, with 
or without frontal and/or occipital compres- 
sion, made its appearance in late prehis- 
toric times throughout a large part of Mexi- 
co. Having seen numerous skulls with lamb- 
doid deformity from other late sites in this 
region, I have no doubt that this is the true 
and significant form of the Sacrificios de- 
formity type. 

The so-called lobation (double or triple) 
which is present in extreme cases of the de- 
formity and which Gosse stressed, appears 

to be a secondary feature. Owing perhaps 
to the pattern of cranial growth under arti- 
ficial restraint there is present sometimes a 
postcoronal depression and a depression 
along the sagittal suture. The opposite con- 
dition, an expansion, usually characterizes 
the midparietal parts in these cases. Thus 
the effect is that of lobation. Because some 
lobation can be found in all classes of de- 
formity and wherever deformity was prac- 
ticed, I attribute it to altered growth proc- 
esses rather than to direct pressure from 
longitudinal and transverse constricting 

In view of what is here pointed out, gen- 
eralizations regarding the distribution of 
deformity types, such as that of Imbelloni, 
are misleading where they have to rely 
upon early descriptions of atypical speci- 
mens and where the chronological sequences 
have not been worked out. 


Dingwall, Eric John. Artificial cranial de- 
formation; a contribution to the study of 
ethnic mutilations xvi+313, pp. London, 

Du Solier, Wilfrido. A reconnaissance on 
Isla de Sacrificios, Veracruz, Mexico. 
Notes on Middle American Archaeology 
and Ethnology, Carnegie Inst. Washing- 
ton, Div. Hist. Res., no. 14: 63-80 (mimeo- 
graphed). 1943. 

Gosse, L.-A. Essai sur les deformations arti- 
ficielles du crane. Ann. Hygiene Publ. 
et Med. Legale, ser. 2, 3: 3 17-393; 4: 1- 
83. (Also published separately in Paris, 
159 pp.+7 pis.) 1855. 

■. Presentation d'un crane deforme de 

Nahoa trouve dans la vallee de Ghovel 
{Mexique). Bull. Soc. Anthrop. Paris 2: 
567-577. 1861. 

Hamy, E.-T. Anthropologic du Mexique. Mis- 
sion scientifique au Mexique et dans VAmer- 
ique Centrale. Recherches zoologiques pub- 
liees sous la direction de M. Milne Ed- 
wards. Pt. 1, 148 pp.+pls. 1-17, 19-21. 
Paris, 1884-1891. 

Imbelloni, J. America; cuartel general de las 
deformaciones craneanas. Actas XXV e 
Congr. Internac. Americanistas (La Plata, 
1932) 1: 59-68. Buenos Aires, 1934. 

Larousse, Pierre. Grand dictionnaire uni- 
versel du XIX e siecle, vol. 3. Paris, 1875. 

Mayer, Brantz. Mexico as it was and as it is, 
xii+390 pp. New York, 1844. 

Serres, A.-E.-R.-A. Note sur deux Micro- 
cephales vivants, attribues a une race amer- 
icaine. Compt. Rend. Hebdomadaires 
Seances Acad. Sci. Paris 41: 43-47. 1855. 



Stewart, T. D. A new type of artificial cranial 
deformation from Florida. Journ. Washing- 
ton Acad. Sci. 29: 460-465. 1939. 

. The cultural significance of lambdoid 

deformity in Mexico. Amer. Journ. Phys. 
Anthrop., n.s., 5: 233-234, abst. 10. 1947. 

. Distribution of the type of cranial de- 

formity originally described under the name 
"tete trilobee." El Occidente de Mexico. 
4 ft Reunion de Mesa Redonda sobre 
Problemas antropol6gicos de Mexico y 
Centro America, 23 a 28 de septiembre 
de 1946, Sociedad Mexicana de Antro- 
pologia. (In pr< 

BOTANY. — Studies in Lonchocarpus and related genera, III : Humboldtiella and 
Callistylon. 1 Frederick J. Hermann, U. S. Department of Agriculture. 

Next to Willardia the two plants most 
frequently confused with Lonchocarpus in 
America are Humboldtiella ferruginea 
(HBK.) Harms and Callistylon arboreum 
(Grieseb.) Pittier. Belonging as they do to 
the tribe Galegeae, both of these shrubs or 
small trees of arid open or sparsely wooded 
slopes, coasts, and roadside thickets are 
readily separable when bearing mature 
fruit from Lonchocarpus and other genera 
of the Dalbergieae by their dehiscent pods. 
Their alternate (occasionally subopposite) 
leaflets likewise distinguish them from 
Lonchocarpus. But, like many tropical 
shrubs, the}^ have a tendency to drop both 
pods and leaves toward the end of the dry 
season and to burst into flower with the 
first rains, generally before the new leaf- 
buds have begun to expand. This leafless 
floAvering stage is particularly attractive to 
collectors, and the superficial resemblance 
between their inflorescences and those of 
some of the Lonchocarpi accounts for the 
large number of specimens in herbaria re- 
ferred at least tentatively to the latter 
genus. In this leafless condition both plants 
seem to be most readily separable from 
Lonchocarpus by their persistent, indurated, 
awl-shaped stipules, by the lack of evident 
bractlets, and by the conspicuous articula- 
tion of the pedicels at both base and apex. 
The calyx, with its deep tube, its compara- 
tively long teeth, and its subbilabiate form, 
so pronounced in the bud, might also be 
considered diagnostic. 

The combination of the half-dozen or so 
characters, for the most part of fundamen- 
tal taxonomic significance, here listed for 
setting off these two plants from Loncho- 
carpus is found in both the monotypic 
genera Humboldtiella and Callistylon. Such 

1 Received August 7, 1947. 

an apparently fortuitous combination of 
characters occurring simultaneously in sup- 
posedly separate genera naturally raises the 
question of their status, and an examination 
of their history reveals that in each case 
this has had a varied career. 

First known of the two was Humbold- 
tiella, originally described by Kunth in 
Humboldt and Bonpland's Nova genera et 
species plantarum (6: 395. 1823) as Robinia 
(?) ferruginea. The basis of Kunth's de- 
scription was a fragmentary flowering speci- 
men, lacking leaves and fruit, collected by 
Humboldt or Bonpland in the Quebrada 
de Tacagua near Caracas, Venezuela, and 
deposited in the quondam Berlin Herbar- 
ium (Field Museum photograph no. 2079). 
In 1924 Harms (Fedde Repert. Spec. Nov. 
19: 12-14) pointed out its departure from 
Robinia, in its rostrate carina and narrow 
pod with unmargined lower suture, and 
from Coursetia, which does possess these 
characteristics, in its broad, short calyx- 
teeth and its coalesced vexillar stamen. 
Because of these discrepancies he proposed 
for it the new genus Humboldtiella. Pittier, 
however, maintained (Journ. Washington 
Acad. Sci. 18 (8) : 209. 1928) that the only 
character that would exclude it from Ro- 
binia "is the absence of a conspicuous 
margin on the upper suture of the pod, 
which would certainly not be sufficient to 
establish a new genus" — a conclusion diffi- 
cult to reconcile with the fact that the plant 
differs from Robinia in possessing monodel- 
phous stamens, a thick, flesh}^ calyx Avith 
short, relatiA r ely blunt teeth, and articulate 
pedicels, in addition to the rostrate keel 
noted by Harms. At any rate, Pittier 
seems to have abandoned his original Ioav 
opinion of the merits of Humboldtiella for 
generic recognition, since in his latest Avork 
on the subject (Leguminosas de Venezuela, 

Feb. 15, 1948 



I — Papilionaceas. Boletin Tecnico No. 5, 
Ministerio Agric. y Cria, Caracas, 1944) we 
find (p. 152) : "Como lo habia hecho entrever 
el Dr. Harms, hace algunos afios, la planta 
que acabamos de definir no pertenece real- 
mente al genero Robinia. Difiere principal- 
mente en la forma del caliz y de las semillas, 
en la ausencia de un ala en la sutura su- 
perior de la legumbre y en otros detalles." 
The second plant, the Colombian "Ra- 
moncillo," was originally described as Cour- 
setia arborea by Grisebach (Flora of British 
West Indies, 183) in 1859. Harms considered 
it, or at least such material of it as was 
available to him, to be sufficiently close to 
Robinia ferruginea to be relegated to its 
synonymy in his transfer of the latter to 
Humboldtiella. This fact seems to have been 
overlooked by Pittier when he expressed 
surprise at finding Coursetia arborea among 
the numbers cited by Harms under his 
Humboldtiella ferruginea (I.e., 208-209). 
His own comment ("Later, when revising 
the Papilionatae for the Venezuelan Her- 
barium, I was surprised to find that, among 
the numbers cited by Dr. Harms as belong- 
ing to his Humboldtiella ferruginea, one 
(no. 5780) apparently corresponds to 
Gliricidia sepium HBK., another (no. 
9078) is unmistakably Coursetia arborea 
Griseb., and only one (no. 6004) belongs to 
the real Robinia ferruginea HBK . . . ") 
seems to be a more fitting basis for surprise 
not only because one would logically expect 
to find Coursetia arborea (Pittier 9078) 
cited under Humboldtiella ferruginea when 
treated as a synonym of that species but 
also because at least the specimens in the 
U. S. National Herbarium of the other 
cited numbers bear out Harms's rather than 
Pittier's determinations, that is, Pittier 
5780 shows no resemblance to Gliricidia 
but is typical Robinia ferruginea, while 
Pittier 6004 is Coursetia arborea, not Robinia 
ferruginea. Following his criticism of 
Harms's work, Pittier proposes segregating 
Coursetia arborea as the type of a new 
genus, Callistylon. In the proposal of 
Callistylon, its distinctiveness from Cour- 
setia is carefully elaborated ; but no reference 
is made to its relationship with Humbold- 
tiella until the necessity of keying out both 
genera arose in the Leguminosas de Vene- 

zuela. Here we find (I.e., 139-140, 152) car- 
ried still further a confusion in the char- 
acters of Humboldtiella and Callistylon, 
which was already apparent in the original 
publication of Callistylon. In the proposal 
of Callistylon the lateral calyx teeth are 
described as more or less acute and the 
lowermost tooth as acute, whereas a new 
description of Robinia ferruginea accom- 
panying the same paper states that the 
calyx teeth in that plant are obtuse. Ex- 
actly the reverse is true, however, as is 
apparent from Kunth's original description 
of Robinia ferruginea (Humboldt & Bon- 
pland 6: 395) as having the calyx "lobis 
. . . acutis" and from an examination of a 
photograph of the type specimen. In the 
generic key to the Venezuelan Papilionatae 
this inversion is perpetuated as "Dientes 
del caliz cortos y todos obtusos ..." for 
Humboldtiella, and "Dientes del caliz 
alargados o desiguales, pero siempre el 
carinal largo y agudo ..." in the lead for 
Callistylon and Coursetia together. The lead 
setting off Callistylon from Coursetia de- 
scribes the calyx teeth as "muy desiguales, 
los 2 superiores obtusos y unidos, los la- 
terales mas cortos y agudos, el inferior otra 
vez mas largo y agudo," but it is only in the 
bud stage that the teeth could strictly be 
called very unequal and the lateral teeth 
shorter than the upper pair. In flowering 
and fruiting material of typical Callistylon 
arboreum the calyx teeth are subequal, the 
lateral and lower ovate and obtuse to sem- 
iorbicular rather than deltoid and acute as 
in typical Humboldtiella ferruginea. The 
original description of Callistylon calls for a 
plant with leaflets "petiolulatis . . . ex- 
stipellatis," whereas the petiolules are ac- 
tually stipellate, and describes the style as 
glabrous as opposed to "basi glabro apicem 
versus utrinque villosulo," whereas the 
villosulous character applies equally well 
to the styles of both plants. Callistylon is 
originally described as having "bracteolae 
parvae et inconspicuae," but the generic 
description in Pittier's later treatment 
specifies "bracteas y bracteolas nullas." 
Diligent search of the 38 sheets of Callisty- 
lon available to the writer has failed to 
reveal any trace of bractlets, but the bracts, 
although caducous, are fully as conspicuous 



VOL. 38, NO. 2 

during most of the bud stage as are the 
equally evanescent bracts of Humboldtiella 
jerruginea. Callistylon is keyed out from 
Coursetia in Leguminosas de Venezuela by 
the additional character of having "valvas 
de la legumbre de dehiscencia plana y no 
elastica," yet the elastic dehiscence of the 
pods in the Pittier collections of Callistylon 
numbered 7833, 10780, and llltf (the 
last cited under the original description) 
could not be more striking. 

There can scarcely be any question that 
the peculiarities of the calyx alone in 
Callistylon are sufficient to set it off generi- 
cally from Coursetia, just as the calyx 
morphology in Humboldtiella is adequate 
basis for removing H. jerruginea from 
Robinia. But between Humboldtiella and 
Callistylon there appears to be no distin- 
guishing character of generic rank. The 
treatment of the two plants by Harms as 
specifically indistinguishable, and Pittier's 
inability to consistently separate the two as 
evidenced by his citation of his collection 
6004 (typical Callistylon arborea) under 
Robinia jerruginea, would seem to be signif- 
icant corroboration of this view. It is 
therefore proposed that the genus Callisty- 
lon be abandoned, and that Coursetia 
arborea be transferred as follows : 
Humboldtiella arborea (Griseb.) Hermann, 

comb. nov. 
Coursetia arborea Griseb., Fl. Brit. W. Ind. 183. 

Callistylon arboreum (Griseb.) Pittier, Journ. 

Washington Acad. Sci. 18 (8): 212. 1928. 

The submergence of this species in Hum- 
boldtiella jerruginea by Harms is not difficult 
to understand when one considers the prev- 
alence of collections which are annoyingly 
intermediate between the two. Of the 53 col- 
lections of Humboldtiella in the U. S. National 
Herbarium, 7 can be conscientiously referred 
to H. jerruginea, 38 may be with varjdng de- 
grees of confidence assigned to H. arborea, but 
8 are too nicely balanced between the two to 
justify outright committal in favor of one or 
the other. On the other hand, the differences 
between the typical forms of both plants (well 
illustrated in plates 80 and 81 in Pittier's 
Leguminosas de Venezuela) are too fundamental 
to permit of their being interpreted as mere 
ecological extremes of a single polymorphic 
species, nor is there a geographic segregation of 

the two sufficiently marked to warrant their 
designation as varieties (i.e., "subspecies," 
as currently employed with increasing fre- 
quency). A hypothesis of hybridization be- 
tween the two species, and that of a contem- 
poraneous, fertile and vigorous type, would 
seem to be the most plausible explanation of 
the frequency of these transitional forms and 
the comparative impurity of many of the forms 
identifiable as possessing predominantly the 
characteristics of one of the extremes. Geo- 
graphically the specimens of H. arborea at 
hand represent localities scattered throughout 
most of northern South America north of 
latitude 5°, from the Department of Bolivar, 
Colombia, eastward to Trinidad and the Ka- 
nuku Mountains in British Guiana; H. jer- 
ruginea is almost restricted to the Venezuelan 
North-Central States of Yaracuy, Carabobo, 
Aragua, Distrito Federal and Miranda, with 
one outlying station in Sucre; and the inter- 
mediates are apparently centered in two dis- 
junct areas, in the Venezuelan North-Central 
States where the ranges of the putative parents 
overlap and on Trinidad and the extreme east- 
ern tip of the Paria Peninsula in Sucre, Vene- 
zuela. The nearest authenticated specimen of 
H. jerruginea is from a locality approximately 
80 miles west of this second area of extensive 
hybridization, although Pittier cites two 
Trinidad specimens that are not available 
for verification. Harms cites a Brazilian col- 
lection under his H. jerruginea, as well as re- 
ports for Panama and Guiana, but his concept, 
of course, included both species. 

The following key is an attempt to empha- 
size the more constant characters by which the 
two species may be separated. Pronounced 
departures from these distinctions are indi- 
cated in the citations of collections below. 

Calyx slightly if at all gibbous, generally tapering 
toward base, averaging 6 mm in length, densely 
ferruginous-strigose, tube l\to2\ times length of 
lateral teeth; calyx-teeth typically deltoid, acute, 
lateral averaging 2 mm long; leaflets generally 
elliptic-lanceolate to oblong-lanceolate, broad- 
est below middle, acute, sparsely strigose above, 
pilose-strigose beneath H. Jerruginea 

Calyx typically gibbous on vexillar side, generally 
rounded at base, averaging 5 mm in length, 
with thinner and paler pubescence, tube 3\ to 4 
times length of lateral teeth; calyx-teeth typically 
ovate to semiorbicular, obtuse, lateral averag- 
ing 1 mm long; leaflets generally oval-elliptic 
to oblong, broadest at middle, obtuse, typi- 
cally glabrous H. arborea 

Feb. 15, 1948 clark: some interesting starfishes and brittle-stars 



(All in U. S. National Herbarium, except Steyer- 

mark collections in Chicago Natural 

History Museum) 


Venezuela: LI. Williams 12331, La Entrada, 
Carabobo (typical); Pittier 7594, between El 
Encanton and Los Teques, Aragua (typical); 
Pittier 9159, between La Victoria and Los Teques, 
Aragua (calyx-lobes mostly obtuse); Pittier 5780, 
La Trinidad de Maracay, Aragua; F. Tamayo 
1305, near Guayas, Aragua; Pittier 11956, Los 
Moriches, Miranda; A. Allart 283, near Las 
Moriches, Miranda; A. Allart 283, near Las Mos- 
tazas, Miranda (calyx pale); Steyermark 62403, 
between Cumanacoa and Cocollar, Sucre (calyx- 
tube nearly 4 times the length of lateral teeth). 


Colombia: R. D. Metcalf 30017, between Me- 
dillin and Antioquia, Antioquia; Dugand & 
Jaramillo 2849, between Cartegena and Turbaco, 
Bolivar; H. M. Curran s. n., Apr.-May 1916, 
Island of Mompos, Bolivar; Bro. Elias 1192, 
Puerto Colombia, Atlantico; Dugand 1181 & 323, 
Barranquilla, Atlantico; Dugand & Jaramillo 
4042, between Juanmina and Cuatrobocas, Atlan- 
tico; Dugand 3640, Juanmina, Atlantico; H. H. 
Smith 935, Santa Marta, Magdalena (calyx-lobes 

Venezuela: Pittier 10780, near Valera, Tru- 
jillo; Pittier 13125, Escuque, Trujillo; F. Tamayo 
1694, Valera, Trujillo; Saer 230, La Ruesga, 
Lara; Saer 247, Barquisimeto, Lara (indument 
ferrugineous) ; A. John 1197, Humocaro, Lara; 
Pittier 7665, between Valencia and Puerto Ca- 
bello, Carabobo (typical); Pittier 9078, between 
Puerto Cabello and San Filipe, Carabobo (calyx- 
lobes unusually acute); Pittier 9413, same, in 
fruit (calyx-lobes blunt); Pittier 7631, near Va- 
lencia, Carabobo; Pittier 10310, between Caracas 
and La Guaira, Dist. Federal; Killip & Tamayo 
37053, Santa Lucia, Miranda; Pittier 600 4, 

Siquire Valley, Miranda (typical); Piiiier 7833, 
Guatire, Miranda; Pittier 1144?, El Sombrero, 
Guarico; Archer 3025, between El Sombrero and 
La Democracia, Guarico (typical); LI. Williams 
12566, El Cristo, Bolivar (calyx acute at base); 
Steyermark 61 488, Bergantin, Anzoategui; Brown, 
Gillin & Bond 21, Paria Peninsula, Sucre; Broad- 
way 796, Cristobal Colon, Sucre; Broadway 809, 
same (typical). 

Trinidad: Britton 4J8, Teteran Bay; Britton 
485, same (leaflets broadest below middle); Brit- 
ton, Hazen, & Mendelson 523, Patos Island; 
Britton, Freeman, & Watts 2704, Chacachacare; 
A. Fendler in 1877-1880, without definite locality; 
Britton & Broadway 466, Lady Chancellor Road. 

British Guiana: A. C. Smith 3092, Kanuku 


Venezuela: Pittier 12601, San Pedro, Yaracuy 
(calyx nearest H. arborea; leaflets nearest H. 
ferruginea); Pittier 8196, between Puerto Cabello 
and San Esteban, Carabobo (calyx proportions of 
H. arborea, but lateral teeth rounded-deltoid; 
indument dense, dark); LI. Williams 10400, 
Carmen, Aragua (calyx gibbous but tapering at 
base, densely hairy, the tube only twice the length 
of the acute lateral lobes; leaflets of H. arborea); 
Broadway 266, Cristobal Colon, Sucre (calyx pro- 
portions of H. arborea but teeth nearly deltoid, 
varying to broadly obtuse; leaflet shape and size 
of H. ferruginea, but glabrous); Broadway 813, 
same (calyx proportions of H. ferruginea and lat- 
eral lobes acute, but vesture pale; leaflets blunt, 
not broadest below middle). 

Trinidad: Broadway s. n., March 7, 1930 (lat- 
eral calyx-teeth deltoid but the tube 4 times their 
length; indument pale); Broadway 8619, without 
definite locality (calyx proportions intermediate, 
the tube 3 times the length of the lateral teeth; 
leaflets blunt, but broadest below middle); 
N. L. & E. G. Britton 2201 coastal thicket (calyx 
teeth deltoid but tube Z\ times as long). 

ZOOLOGY.- — Some interesting starfishes and brittle-stars dredged by the Atlantis 
in the mid- Atlantic. 1 Austin H. Clark, U. S. National Museum. 

The Woods Hole Oceanographic Institu- 
tion has recently submitted to me for ex- 
amination, through Dr. Louis W. Hutchins, 

a small but interesting collection of star- 
fishes and brittle-stars dredged by the At- 
lantis in the mid- Atlantic. The expedition 
on which these were found, Cruise Xo. 150 
of the Atlantis, was sponsored by the Na- 
tional Geographic Society, Columbia Uni- 

1 Contribution from the Woods Hole Oceano- 
graphic Institution No. 410. Published with the 
permission of the Secretary of the Smithsonian 
Institution. Received November 5, 1947. 

versity, and the Woods Hole Oceanographic 
Institution. I am much indebted to the 
Institution and to Dr. Hutchins for the 
privilege of studying this collection. 

The species represented in the collection 
are the following: 

Hyphalaster parfaiti E. Perrier 
Locality. — Atlantis station 15; mid-Atlantic 

west of Gibraltar (lat. 35° 37' N., long. 30° 51' 

W.); 3,200 meters; August 16, 1947. Seven 


Notes. — The details of the seven specimens 

are as follows : 








orga n 



34 mm. 

14 mrn. 

































Pythonaster atlantidis n. sp. 

Description. — # = 165 mm, r=28 mm. The 
height at the center of the abactinal system is 
26 mm. The rays are 31 mm wide at the base, 
tapering rapidly to 10 mm at 30 mm from the 
base, from this point tapering gradually to 3 
mm at the tip. A shallow interradial sulcus 
runs from the middle of each interradius to the 
center of the disk. The animal may be de- 
scribed as broadly stellate, with each point of 
the star produced into a long and slender arm. 

There is no abactinal skeleton, but the skin 
is filled with minute rounded plates, which are 
almost contiguous on the disk and arm bases 
but become scattered on the outer part of the 
arms. Among these are the much larger and 
denser circular lenticular plates, which carry 
the groups of spines. 

On the midline of each ray, running from the 
center of the disk to the end of the swollen 
portion of the arm bases, is a band 10-15 mm 
wide consisting of groups of from 2 or 3 to 
about 10, commonly 5 or 6, very slender and 
delicate spines 2-3 mm long arising from a 
common base on a deeply embedded lenticular 
plate, the whole enclosed in a bag of thick skin 
by which the spines are entirely concealed. In 
a patch about 15 mm long on the inner part of 
the disk these spine-containing sacculi are in- 
termixed with papulae. On the disk this band 
of sacculi is bordered on each side by a patch of 
bare skin through which the large eggs, 2 mm 
in diameter, are visible. These bare patches 
may bear two or three sacculi including 4-7 
spines. The apical region is closed by 5 large 
radially placed triangular flaps consisting of 
numerous delicate spines enclosed in a web of 
thick skin. The madreporite is situated at the 
inner end of an interradial sulcus at the outer 
end of the line between two of these valves or 
flaps. Just beyond the central valves and bor- 
dering the interradial sulcus for about half its 
length are a few cup-shaped structures consist- 
ing of about a dozen slender spines radiating 
upward from a common base and enclosed in 
thick skin. 

The arms beyond the swollen basal portion 
are in cross section triangular with the apex 
rounded, about as high as broad, becoming 
higher than broad toward the arm tips. 

The adambulacral plates have a strongly 
curved crest which is set at an angle of about 
60° with the axis of the arm, the distal end of 
each crest being considerably farther inward 
than the proximal end of the crest following. 
These crests carry a row of 8-10 slender spines 
which are closely placed with their swollen 
bases contiguous and are united by a web; 
the innermost spines are about 1 mm long, 
those following gradually increasing in length 
to the outermost, which is 3 mm long. The 
outer part of each adambulacral plate bears a 
single much stouter spine 4 mm long, these 
stouter spines supporting a broad fin-like web 
which runs along the actinolateral border of 
the arm resembling the actinal web of many 
Pterasteridae. Beyond the proximal swollen 
portion of the arms the adambulacral spines be- 
come reduced in number, being usually 5. 

Running upward from each adambulacral 
plate to the dorsolateral border is a band about 
1.5 mm broad composed of numerous fine 
spines arising in small groups, sometimes 
singly, from a row of independent concretions, 
the whole band being enclosed in a continuous 
envelope of thick skin. Beyond the swollen 
proximal portion of the ray the bands of the 
two sides meet in the middorsal line. Distally 
these bands become narrower, the concretions 
bearing only 1-3 spinelets. 

The pair of mouth plates is 9 mm broad at 
the mouth edge, 2 mm broad on the outer bor- 
der, and 7 mm long. The line of union of the 
two plates is raised into a rounded crest. The 
inner border at the mouth edge is everted and 
curved, and the outer borders are somewhat 
concave. The everted inner border of each 
mouth plate bears 4 large spines, beyond 
which are 5 much smaller spines. There are no 
spines on the actinal surface. The mouth 
plates recall those of some of the Pterasteridae, 
as for instance Hymenaster perissonotus. 

The mouth is circular, 23 mm in diameter. 
The very large stomach is empty. The tube 
feet, in two series, are large and stout and end 
in a large sucking disk. 

On the arms beyond the swollen basal por- 
tion the ambulacral plates are long and slender. 
The sides of the lower half diverge gradually 
so that the end adjoining the adambulacrals is 

Feb. 15, 1948 clark: some interesting starfishes and brittle-stars 


about 2 mm broad, the central portion of the 
plate being only 0.5 mm wide. The upper third 
of the ambulacrals is abruptly broadened in 
the form of a broad Y with very thick arms. 
When viewed from the exterior the distal arm 
of the Y is entirely concealed by the proximal 
arm of the Y of the ambulacral following, which 
imbricates over it. This causes the ambulacrals 
when viewed from the exterior, to appear ab- 
ruptly bent proximally, but when viewed from 
the interior they are seen to be straight with 
both arms of the Y about equal. 

The adambulacrals as viewed from the ac- 
tinal edge are seen to be narrowly rhombic, 
about 3 mm long and 0.75 mm wide. The inner 
side of the lower angle of the rhombic figure is 
abruptly swollen, appearing as if a rather thick 
flat pad had been soldered to it. The distal edge 
of this pad is evenly curved and bears the spines 
that form the adambulacral comb. The stout 
adambulacral spine is situated at about the 
middle, and widest part, of the adambulacral 
plate at some distance from the comb-bearing 
pad and entirely out of line with it. The adam- 
bulacral plates are strongly imbricated so that 
the large spine is in line with the spines of the 
comb of the plate following, to which it appears 
to belong unless the arm is cleaned, when its 
true relationship becomes apparent. 

Just above each adambulacral and parallel 
to its long axis, lying practically on its surface, 
there is a long, slender and delicate plate 4 or 
5 times as long as broad, and a little over half 
as long as the adambulacral; in some cases 
there are two of these lying closely side by side. 
These plates appear to be vestigial marginals. 

In a line from these vestigial marginals to 
the midradial line, as viewed from the inner 
surface of the skin, there is a regular row of 
elongate sharply carinate plates 2 or 3 times as 
long as broad closely placed with their long 
axes parallel. Between these regular columns 
of elongate plates (from the center of which on 
the outer side spines arise) there are very 
numerous and closely packed minute concre- 

Distally the plates in these columns become 
smaller, fewer, and more widely spaced, and the 
concretions become more widely scattered. 

Locality. — Atlantis station 15; mid- Atlantic 
west of Gibraltar (lat. 35° 37' N., long. 30° 51' 
W.); 3,200 meters; August 16, 1947. One speci- 
men (type, U.S.N.M. No. E. 7175). 

Notes. — The genus Pythonaster, the sole 

genus in the family Pythonasteridae, includes 
only the type species Pythonaster murrayi 
Sladen described from a single specimen 
dredged by the Challenger at station 323 east 
of Buenos Aires, Argentina, in 1,900 fathoms, 
on February 28, 1876. 

The type specimen of P. murrayi is slightly 
smaller than the type specimen of P. atlan- 
tidis, and the bases of the rays are much less 
swollen; but this difference may be due to the 
occurrence of large masses of eggs in the type 
of B. atlantidis. 

According to Sladen's description and figure 
the actinolateral areas of P. murrayi are tra- 
versed by regular rows of isolated skin sacks, 
whereas in P. atlantidis these skin sacks are 
coalesced into continuous lines. The skin sacks 
on the abactinal surface of P. murrayi are much 
smaller and more widely separated than those 
of P. atlantidis in which they are almost or 
quite in contact. 

In P. murrayi the adambulacral combs are 
composed of about 6 spines the number be- 
coming reduced to 5 distally. In P. atlantidis 
there are 8-10 proximally, 5 distally. In P. mur- 
rayi the outermost and largest spine on the 
adambulacrals is counted as one of the comb, 
being united to the next by a web. In P. at- 
lantidis the outermost spine on the adambulac- 
rals is abruptly larger than the others and rela- 
tively much larger than in P. murrayi and is 
joined to the outermost spines on each side by 
a web, all the spines together and the web 
forming a continuous broad actinolateral web 
or fin which appears to be absent in P. mur- 

The mouth plates of P. murrayi as figured 
are not of the same shape as those of P. at- 
lantidis. They bear 3 or 4 large inner mouth 
spines and 3 small outer ones instead of 4 large 
and 5 small as in P. atlantidis. Furthermore, 
there are in P. murrayi spines on the actinal 
surface of the mouth plates which are not pres- 
ent in P. atlantidis; however, in the latter these 
may have been rubbed off during capture. 

With only a single specimen of each species 
it is impossible to form any idea of the limits 
of variation. It is possible, though not probable, 
that P. atlantidis is a more fully developed ex- 
ample of the species represented by P. mur- 
rayi. It is also possible, though not probable, 
that P. atlantidis is a female of the species of 
which the type of P. murrayi is the male; but 
Sladen did not give the sex of his specimen. 



Asteroschema inornatum Koehler 

Locality. — Atlantis station 6; mid-Atlantic 
east of Bermuda (lat. 30° 06' N., long. 42° 08' 
W.), 1,554 meters; August 8, 1947. One speci- 

Notes. — As the original description of this 
species was brief and lacking in detail the fol- 
lowing supplementary information derived 
from the present specimen may be of interest: 

The disk is 6 mm in diameter, stellate with 
truncated angles and regularly incurved inter- 
radial borders; the outer ends of the radial 
shields are raised about 0.75 mm above the 
arm bases. The arms are about 60 mm long, 
slender, 1.3 mm broad at the base and tapering 
gradually to a delicate tip; they are only 
slightly higher than broad. 

The disk is covered rather thickly with mi- 
nute granules of coarse and spongy structure. 
The radial shields are rather broad, widest in 
the middle, the outer border having a broadly 
obtuse median angle, and those of each pair are 
almost in contact. They extend inward for 
about two-thirds distance to center of disk. 

The first two tentacle pores have no scales. 
The next five have a single tentacle scale. Those 
following have two scales of which the large 
inner scale is about a segment and a half in 
length. At the arm tip the outer scale is only 
slightly longer than the inner, but neither as- 
sumes a hooklike shape. The larger inner scale 
is cylindrical in the basal third or fourth, from 
that point being less dense and tapering to the 
tip. The distal three-fourths bears on the outer 
side numerous long and slender sharp spine- 
lets, longest near the tip, which through the 
dried skin appear as small conical points. There 
are similar but smaller spinelets on the inner 
side. The small outer scale is similarly armed. 

There are 5 stout triangular teeth. 

On the sides of the jaw plates within the 
mouth there are about 8 rather large and well 
separated granules; these are irregularly scat- 
tered, the group as a whole running upward and 
outward, the outermost being a pair, one on 
each side of the uppermost tooth. 

The skin of the arms contains large delicate 
filmy plates that appear to form an almost 
continuous pavement. 

The color in alcohol is pale dull pinkish, be- 
coming dark purple on the outer half or more 
of the tentacle scales. 

This species was described by Prof. Rene 

Koehler (1907) on the basis of two specimens 
dredged by the Talisman in the Gulf of Gas- 
cony (the inner part of the Bay of Biscay) (lat. 
45° 59' N., long. 6° 29' W.) in 1,480 meters, 
coral bottom, on August 30, 1883. Sanderson 
Smith (1888) gives this station as No. 156. 
Koehler gave the color in alcohol as white. 
A single specimen was subsequently dredged 
by the Princesse- Alice at station 2248, in lat. 
37° 02' 30" N., long. 27° 35' W., in 1,478 meters, 
on September 6, 1905 (Koehler, 1909). In his 
report on the Princesse- Alice echinoderms 
Koehler did not amplify his original descrip- 
tion, but he gave a colored figure showing the 
animal as deep pink, lighter on the radial 
shields and becoming lighter on the arms after 
the basal 20 mm. 

Ophiura inornata (Lyman) 

Locality. — Atlantis station 15; mid-Atlantic 
west of Gibraltar (lat. 35° 37' N., long. 30° 51' 
WO; 3,200 meters; August 16, 1947. Seven 

Notes. — In these specimens the oral shields 
are in most divided into two elongate plates 
meeting in the median line as an angle of about 
90°. The jaw plates, the adoral shields, and the 
two sections of the oral shields are similar, and 
of about the same size, appearing as three 
similar chevrons. 

In a small specimen one of the oral shields 
is undivided. In other cases it appears as if the 
third or outer chevron was composed in reality 
of a pair of supplementary plates intercalated 
between the oral shield, much reduced in size, 
and the adoral shields. 

Opiomusium armigerum Lyman 

Locality. — Atlantis station 15; mid-Atlantic 
west of Gibraltar (lat. 35° 37' N., long. 30° 51' 
W.); 3,200 meters; August 16, 1947. Two spec- 


Koehler, Rene. Descriptions des ophiures nou- 
velles recueillies par le Travailleur et le Talis- 
man pendant les campagnes de 1880, 1881, 
1882, & 1883. Mem. Soc. Zool. France 19: 
(for 1906): 30, pi. 3, figs. 46, 47. 1907. 

. Resultats des campagnes scientifioue ac- 

complies sur son yacht par Albert l er , Prince 
souverain de Monaco, fasc. 34: 205, pi. 7, 
fig. 1. 1909. 

Smith, Sanderson. Lists of dredging stations in 
North American waters from 1867 to 1887. 
Ann. Rep. Commissioner Fish and Fisheries 
for 1886: 871-1016 (p. 984). 1888. 

Feb. 15, 1948 

proceedings: the academy 




The 414th meeting of the Board of Man- 
agers, held in the Cosmos Club, January 12, 
1948, was called to order at 8:05 p.m. by the 
President, Dr. Waldo L. Schmitt. Others 
present were: H. S. Rappleye, N. R. Smith, 
H. A. Rehder, M. A. McCall, H. B. Collins, 
Jr., J. S. Wade, A. Wetmore, W. W. Rubey, 
L. E. Yocum, W. A. Dayton, C. A. Betts, 
A. 0. Foster, C. L. Gazin, and, by invitation, 
G. P. Walton, J. I. Hoffman, E. A. Chapin, 
and A. T. McPherson. 

The Chairman of the Committee on Mono- 
graphs, Dr. E. A. Chapin, announced that 
estimates had been obtained from several 
printers for the publication of Dr. Herbert 
Friedmann's monograph The Parasitic Birds 
of Africa. These ranged from about $1,500 to 
$2,500 for 1,000 paper-bound copies. Buckram 
binding increased the estimates $450 to $460. 
From the record of the sale of Dr. Friedmann's 
earlier monograph of the cowbirds, Dr. Chapin 
considered that the present volume, published 
possibly as a supplement to the Journal, 
would in all probability pay its way and 
recommended that the Academy accept it. 
The monograph was referred to the Board of 
Editors for further review and recommenda- 

Ten persons were elected to membership. 

The Chairman of the Committee on Science 
Legislation, Dr. A. T. McPherson, briefly 
summarized current activity in science legisla- 
tion, on which he would report more fully 
at the annual meeting. His talk was concerned 
principally with the progress of the Foreign 
Scholarships program and on H. R. 3342. Dr. 
McPherson proposed that serious considera- 
tion be given to the formation of a study group 
on science legislation. By an informal showing 
of hands the outgoing Board recommended to 
the incoming Board that a Science Legislation 
Committee or Study Group be reappointed for 
the coming year. 

The Acting Chairman of the Committee on 
Academy Awards for Scientific Achievement in 
1947, Dr. Henry B. Collins, Jr., presented 
the following recommendations for the three 

Biological Sciences: None. 

Engineering Sciences: Harry Warren Wells, 

of the Department of Terrestrial Magnetism of 
the Carnegie Institution of Washington, in 
recognition of his distinguished upper-air re- 
search and organization of a world-wide network 
of ionospheric stations. 

Physical Sciences : Dr. Robert D. Htjntoon, of 
the National Bureau of Standards, in recognition 
of his distinguished service in the advancement 
of electronics and its applications to other sci- 
ences and to modern ordnance. 

The Board accepted the recommendations, 
but during the discussion it developed that 
there existed some misunderstanding as to the 
age limit of award winners. The Board voted 
that for future purposes the nominee for the 
award must not have passed his 40th birthday 
during the year for which the award was made. 

The Chairman of the Committee on the 
Journal and its Improvement, Dr. R. J. 
Seeger, reported as follows: 

The Committee has had several meetings, in- 
cluding those of special subcommittees, over a 
period of almost a year. During this time it has 
reviewed the reports of similar previous com- 
mittees; it has also consulted with various past 
Editors of the Journal and with the George 
Banta Publishing Co. After considering many as- 
pects the Committee unanimously makes the 
following recommendations pertaining to the 
Journal and its improvement: 

(1) That an index of the Journal be prepared 

and published for the first 40 volumes 
and for each succeeding 10 volumes. 

(2) That the address of the Retiring President 

of the Academy be published without 
charge, including the first 50 reprints 
with covers. 

(3) That each author be allowed, without 

charge, the equivalent of the first 50 
four-page reprints without covers. 

(4) That each invited Academy speaker whose 

address is published in the Journal be 
allowed, without charge, the first 50 re- 
prints without covers. 

(5) That the size of the Journal be increased 

to approximately 500 pages per year and 
that the Journal be published economi- 
cally in issues of 32 or 48 pages — say, 
6 issues of each size. 

(6) That in each issue one page (the Editorial 

Page) be allotted to the President of the 
Academy for his own use or designation. 

(7) That the list of officers and committees 

(standing and special) of the Academy be 



published on the third cover page (inside 
back) of each issue of the Journal. 

(8) That Editors of the Journal be continu- 

ously vigilant in seeking papers of gen- 
eral interest, in securing copies of out- 
standing Academy addresses (footnote 
reference to occasion of delivery), and 
in selecting papers from different fields 
for any particular issue. 

(9) That the Board of Managers authorize the 

appointment of a committee to study the 
functions of the Academy and to formu- 
late a program that will integrate these 
functions, including the Journal. 

The Board voted that the recommendations 
of the Journal Committee be brought up at 
the next meeting of the Board. 

The Secretary, Dr. C. L. Gazin, announced 
the death of Maj. Gen. Georges Perrier, 
honorary member, formerly of Paris, France, 
on February 16, 1946; Hardee Chambliss, 
formerly of Catholic University of America, on 
June 1, 1947; and Robert H. Lombard, for- 
merly of the Norton Company, Research 
Laboratories, Worcester, Mass., on October 11, 

Dr. Harvey I. Curtis was transferred to the 
list of retired members, effective December 31, 

The following letter to the Secretary from 
Dr. P. W. Bridgman was read to the Board: 

It is a great pleasure to learn that I have been 
transferred from regular membership to honorary 
membership in the Washington Academy of Sci- 
ences. Will you please express my appreciation to 
the Board of Managers. 

The Treasurer, H. S. Rappleye, announced 
that the auditing committee had met that 
morning, examined his books and the Academy 
securities, and gave him a "clean bill of 
health." He announced also that the transfer 
of investment certificates to an account book 
form in the First Federal Savings and Loan 
Association, as authorized at the 413th meet- 
ing, had been made. 

The President, Dr. W. L. Schmitt, stated 
that expenses of the Meetings Committee had 
exceeded the amount of the budget for this 
purpose and that in order to secure outstanding 
speakers he had paid out $100.30 more than 
allowed. It was his hope that, at the close of 
the Academy Year 1947, the unexpended 
amounts from the budgets of other offices and 
committees might be transferred to that of the 
Meetings Committee, up to the amount of 
$100.30, but without increasing the budget for 
the year. The Board authorized that Dr. 
Schmitt be so reimbursed for the expenditures 
made by him on behalf of the Meetings Com- 

The meeting was adjourned at 9 : 45 p.m. 

C. L. Gazin, Secretary 

Officers of the Washington Academy of Sciences 

President Frederick D. Rossini, National Bureau of Standards 

Secretary C. Lewis Gazin, U. S. National Museum 

Treasurer Howard S. Rappleye, Coast and Geodetic Survey 

Archivist Nathan R. Smith, Plant Industry Station 

Custodian and Subscription Manager of Publications 

H. A. Rehder, U. S. National Museum 

Vice-Presidents Representing the Affiliated Societies: 

Philosophical Society of. Washington Walter Ramberg 

Anthropological Society of Washington T. Dale Stewart 

Biological Society of Washington John W. Aldrich 

Chemical Society of Washington Charles E. White 

Entomological Society of Washington C. F. W. Muesebeck 

National Geographic Society Alexander Wetmore 

Geological Society of Washington William W. Rtjbey 

Medical Society of the District of Columbia Frederick O. Coe 

Columbia Historical Society ; Gilbert Grosvenor 

Botanical Society of Washington Ronald Bamford 

Washington Section, Society of American Foresters William A. Dayton 

Washington Society of Engineers Clifford A. Betts 

Washington Section, American Institute of Electrical Engineers 

Francis B. Silsbee 

Washington Section, American Society of Mechanical Engineers 

* Martin A. Mason 

Helminthological Society of Washington Atjrel O. Foster 

Washington Branch, Society of American Bacteriologists Lore A. Rogers 

Washington Post, Society of American Military Engineers . Clement L. Garner 

Washington Section, Institute of Radio Engineers Herbert Grove Dorsey 

Washington Section, American Society of Civil Engineers Owen B. French 

Elected Members of the Board of Managers : 

To January 1949 Max A. McCall, Waldo L. Schmitt 

To January 1950 F. G. Brickwedde, William W. Diehl 

To January 1951 Francis M. Defandorf, William N. Fenton 

Board of Managers All the above officers plus the Senior Editor 

Board of Editors and Associate Editors [See front cover] 

Executive Committee Frederick D. Rossini (chairman), Walter Ramberg, 

Waldo L. Schmitt, Howard S. Rappleye, C. Lewis Gazin 

Committee on Membership 

Harold E. McComb (chairman), Lewis W. Butz, C. Wythe Cooke, William 

, . . . W. Diehl, Floyd D. Felton, Regina Flannery, George G. Manov 

Committee on Meetings Raymond J. Segeer (chairman), 

Frank P. Cullinan, Fred L. Mohler, Francis O. Rice, Frank Thone 

Committee on Monographs: 

To January 1949 Lewis V. Judson (chairman), Edward A. Chapin 

To January 1950 Roland W. Brown, Harald A. Rehder 

To January 1951 William N. Fenton, Emmett W. Price 

Committee on Awards for Scientific Achievement (Karl F. Herzfeld, general chairman): 

For the Biological Sciences 

C. F. W. Muesebeck (chairman), Harry S. Bernton, Chester W. Emmons, 
Elmer Higgins, Mario Mollari, Gotthold Steiner, L. Edwin Yocum 

For the Engineering Sciences 

Harry Diamond (chairman), Lloyd V. Berkner, Robert C. Duncan, 
Herbert N. Eaton, Arno C. Fieldner, Frank B. Scheetz, W. D. Sutcliffe 

For the Physical Sciences 

Karl F. Herzfeld (chairman), Nathan L. Drake, Lloyd D. Felton, 
Herbert Insley, William J. Rooney, Robert Simha, Michael X. Sullivan 

Committee on Grant-in-aid for Research 

. .F. H. H. Roberts, Jr. (chairman), Anna E. Jenkins, J. Leon Shereshevsky 

Representative on Council of A. A. A. S. S Frank Thone 

Committee of Auditors 

William G. Brombacher (chairman), Harold F. Stimson, Herbert L. Haller 

Committee of Tellers 

. . .John W. McBurney (chairman), Roger C. Bates, William A. Wildhack 


P» e e 

Physics. — The measurement of high voltage. F. M. Defandorf .... 33 

Chemistry. — Standardization of the pH scale. Roger G. Bates and 
Edgar Reynolds Smith 61 

Anthropology. — The true form of the cranial deformity originally de- 
scribed under the name "tete trilob£e." T. D. Stewart 66 

Botany. — Studies in Lonchocarpus and related genera, III: Humbold- 
tiella and Callistylon. Frederick J. Hermann 72 

Zoology. — Some interesting starfishes and brittle-stars dredged by the 
Atlantis in the mid-Atlantic. Austin H. Clark 75 

Proceedings: The Academy 79 

This Journal is Indexed in the International Index to Periodicals 

Vol. 38 

March, 15, 1948 

No. 3 





James I. Hoffman 


Alan Stone 


Frank C. Kracek 



Lawrence A. Wood 


J. P. E. Morrison 


Elbert L. Little, Jr. 


Richard E. Blackwelder 


James S. Williams 


Waldo R. Wedel 


Irl C. Schoonover 




450 Ahnaip St. 

at Menasha, Wisconsin 

Entered as seoond class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a special rate of postage provided for in the Aot of February 28, 1925. 

Authorised January 21, 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author* 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate): 

Copies 4 pp. 8 pp 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7.50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8.00 0.60 

Vol. 23 to current vol. (12 numbers per vol.) 7.50 0.90 

* Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U, S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges, — The Academy does not exchange its publications for those of other 




Vol. 38 

March 15, 1948 

No. 3 

PHILOLOGY. — English-language surnames of biological origin. 1 Perez Simmons, 
Fresno, Calif. (Communicated by Joseph S. Wade.) 

In the normal course of events every per- 
son born in the United States possesses im- 
mediately two permanent designations, his 
family name and his national citizenship. 
To most people these are perhaps equally 
precious, but the first is of course the more 
personal. Family names are treasured inher- 
itances, and even those that may seem, to 
the other fellow, to be trivial, amusing, or 
absurd are likely to be continued for scores 
of generations. Surnames are our handholds 
on the family tree, and most of us refuse to 
let go. 

The surnames of biological origin listed in 
this report I have found in telephone direc- 
tories. Included are plant and animal names 
and terms referring to the parts of plants 
and animals (Daisy, Moose, Seed, Antler). 
Associations' of plants and of animals (Or- 
chard, Herd, Covey) are left out, as are 
also descriptive-anatomical names (Bare- 
foot, Smallback) and physical abnormali- 
ties. Not too strictly, I have omitted va- 
rieties of domesticated plants and animals 
(Pippin, Beagle), some products made from 
and most constituents of plants and ani- 
mals (Cream, Hay, Leather, Pickles, Pitch), 
and obsolete words (Poe and Coe for pea- 
cock and jackdaw). Most compounds such 
as those of big, bird, berry, corn, crow, gold, 
long, short, silver, small, wood, and the 
names of colors are left out (Birdseye, 
Greenleaf, Redapple). No attempt has been 
made to trace derivations. One advantage 
of making my own rules has been freedom 
to appropriate names of dual or multiple 

Books and other references on the subject 
point out that in Biblical times, and until 

1 Received August 22, 1947. 

after the Norman Conquest in 1066, or- 
dinary people had only one name, such as 
Joseph. As populations increased, the con- 
venience of having another name was recog- 
nized, and surnames began to be adopted. 
That was some 700 years before Linnaeus 
established the binomial system for the 
science of biology. 

In Great Britain the cities were ahead of 
the countryside in making the change. 
Whether the surnames established there 
were allotted to families or were chosen by 
them without official guidance is not on 
record in the sources that I have examined. 
It seems likely that both methods were 
used. From the wide coverage of the names 
that I shall list farther on, one would sur- 
mise that some system was involved, at one 
time or another. 

Surnames have been grouped under four 
classifications: locality names derived from 
the home of the bearer (Kent) ; patronym- 
ics, those referring to the father's Chris- 
tian name (Robertson) ; names taken from 
occupations (Weaver); and nicknames 
(Fox, House, Longfellow, Stone). 

On the whole, the family names that I 
have found in my hunting through tele- 
phone directories have a homespun dignity, 
an earthy quality that reflects what was 
prominent and important in day-by-day 
living on the island of Great Britain 25 or 
more generations ago. In those days men, 
women, and children lived closer to their 
farm animals, their vegetable gardens, and 
to wild nature than most people do now. 

As my collection has grown, and with the 
use of the Oxford Dictionary and other 
dictionaries, it has become clear to me that 
there is no need to seek obscure explana- 
tions for the greater number of the biologi- 


mk 2 4 191® 



cal surnames found. Nearly all of them, 
cmite evidently, were taken directly from 
the names of living things of interest in 
everyday life in Britain before the discovery 
of America. Doubled consonants, final e's 
added, and some other departures from 
modern spellings are for the most part 
straightforward legacies from olden times. 

My first adventure into the vast assem- 
blage of American family names was by way 
of the Fresno, Calif., Telephone Directory. 
Since this modest book involved scanning 
only some 200 linear feet of columns — about 
25,000 names — the entire list was searched. 

There were 15 tree names listed in the 
Fresno book: 













And 22 varieties of plants which attract 
attention because of their beauty or utility : 



















General plant terms found were Bush and 
Herb, and plant parts were well represented 
among Fresno surnames, as follows: 
















As might be expected, nearly all of the 
animal names referred to vertebrates. 
Fishes, so important to the island-dwelling 
British, were well represented. All refer to 
European species but all of their names 
have been transplanted for use in North 















Farther up the scale are the reptiles, rep- 
resented by one name only: Asp. 

Birds, being conspicuous and abundant 
as to species, were of frequent occurrence. 
There were 19 of them, and the majority 
are of rather large size: 




















Excluding domesticated forms there were 
only 11 mammals, if Fisher is considered to 
be an occupational surname: 
Badger Fitch Lion 

Batt Fox Puma 

Bear Hare Wolf 

Fawns Hart 

Among miscellaneous terms were the 
following : 

Bird Fish Maiden 


Anatomical terms were plentiful: 


















Names of domestic animals, including 
poultry, number more than 50. There usu- 
ally are separate designations for mature 
males and females, newborn young, juve- 
niles of both sexes, and unsexed males. In 
addition, call-names and pet names have 
been used for centuries. Surnames based on 
these terms probably are nearly all pre- 
Columbian, since the only domesticated 
animals native to the Western Hemisphere 
that have been put to general use by the 
settlers of North America are the turkey 
and the muscovy duck. The dog was com- 
mon to both the New World and the Old at 
the time of the discovery of the New. 

In the Fresno Telephone Directory there 
were four surnames that refer to the names 
of farm animals: 

Barrows Lamb Pigg 


Following the page-by-page scanning of 
the Fresno directory it seemed necessary to 
change the method of hunting for surnames 
of biological origin. A detailed examination 

Mar. 15, 1948 simmons: english-language surnames of biological origin 


of the 455 pages from Aabel to Zzylch in the 
San Francisco book and of the 1,032 pages 
of Los Angeles names from Aaen to Zylstra 
seemed too much of a project. Therefore, 
lists of terms that probably had been 
adopted as surnames were compiled. The 
chief sources of these were the indexes of 
reference books on botany, horticulture, 
and zoology. All the findings from both 
cities were combined and are listed here 
under several classifications. 

Forage plants, small grains, fiber plants, 















plants, etc.: 

















Garden ornamentals, perfume plants: 

Aster Fern Marigold 

Balsam Flagg Pink 

Broom Heather Poppy 

Cosmos Lavender Primrose 

Fruit and ornamental trees were: 









Other botanical terms in the San Fran- 
cisco and Los Angeles telephone books were 
Nut, Plant, Posey, Spore, Stem, and Vine. 

As was the case in the Fresno directory, 
the two leading Pacific coast cities revealed 
an extensive list of fishes : 








There was a single form of amphibian, 
Frogge, and one reptile, Turtle. The birds 
were even more abundant; 27 were found, 
not including Loon, a Chinese name: 










































Domestic animals were represented by 
the following useful inhabitants of the 
farmstead : 































more mammals 

were tropl 

the hunt: 

















The directories of San Francisco and Los 
Angeles provided the following additions to 
the list of anatomical terms: 






























Having had good hunting on the Pacific 
coast, I prepared for larger fields by adding 
to my list of unfound possibilities until it 
numbered about 625. Of much help were 
the many plants and animals illustrated in 
a dictionary, and I included in the list about 
50 biological surnames encountered in gen- 
eral reading. All the latter were found sub- 
sequently in telephone directories of cities 
in the eastern part of the United States. 

Giant among telephone directories is the 
Chicago book, because New York City sub- 



scribers are grouped in several separate 
volumes based on the boroughs, and the 
third telephone city of the world, London, 
divides its directory into . two volumes. 
From Aabad to Zyzik in the Chicago direc- 
tory the distance was 1,741 four-column 
pages, a total column length of 6,021 feet. 
From a systematic standpoint the most 
lowly among the surnames of plant origin in 
Chicago was Kelp, followed by Brake. 
Wild and cultivated plants, except trees, 
were as follows: 






















There were 17 trees and shrubs, as well as 
the name Tree. 


















Names that refer to plant anatomy were : 




Although lists already given include 
many fishes, there were still others in the 
Chicago directory: 










And still 

more birds: 


















Among the following mammals were 
several additions to the list of domestic 
animals : 




















Chicago furnished additions, also, to the 
group of anatomical terms: 





















Following the expedition through the 
Chicago directory, the list of possibilities 
were checked through the 'phone books of 
Boston, Brooklyn, and Detroit. From 
Aaberg to Zynsky, in Boston, was a dis- 
tance of 804 pages and 2,780 linear feet of 
columns; there were 784 pages in the Brook- 
lyn book, from Aa to Zytynski; and the 
Detroit directory, from Aaberg to Zzzpt, 
was 910 pages in length. Trophies from all 
three hunts are combined in the lists which 

In Detroit I found Winston Churchill's 
famous trio of Blood, Sweat, and Tears. 
Other materials found in the three books 
were Butter, Milk, Ivory, and Breath. 

The botanical names were these : 



















There were 

more fishes: 









One reptile, Blackadder, occurred. 

In addition to the many species of birds 
recorded from directories previously exam- 
ined, there w T ere 10 in Boston, Brooklyn, 
and Detroit, with Brood for good measure. 



A few more mammals : 





And further anatomical names: 




Mar. 15, 1948 simmons: english-language surnames of biological origin 


After having explored seven telephone 
directories I began to get the feel of the 
undertaking. A sense of the kind of words 
likely to be found as biological surnames 
had developed, so I went back over some of 
the indexes of botanical and zoological 

More possibilities were obtained by 
scanning references on the botany and zo- 
ology of the British Isles. This was a pleas- 
ant experience because so many of the 
common names used in the mother country 
have been adopted here. On the other hand, 
many were wholly strange. If an American, 
reading about rural England, should find 
this statement: "Then I glimpsed an effet 
as it dashed from a patch of squitch and 
gobbled a mawk beneath a paigle," he 
would be confused. Having translated, he 
would find that a lizard had left its hiding 
place in couch grass (Agropyron repens) and 
eaten a maggot under a cowslip (Primula 
veris). We need to understand our neigh- 
bor's language better! 

Several hundred possibilities resulted 
from this more effective approach. 
Equipped with the new list I hunted again 
through the telephone directories of Los 
Angeles, San Francisco, Chicago, Detroit, 
Boston, and Brooklyn. Using both lists, I 
then examined the Philadelphia Telephone 
Directory. The plants found in the seven 
books, omitting a number of unfamiliar 
ones, were as follows: 














































There were more plant parts: 

























Two of the surnames found were trans- 
fers from the names of amphibians: Asker, 
Eft. Three reptiles, Boa, Luth, and Ser- 
pent e, were discovered. 

Birds and their parts were as follows: 



















and their structures were 




















Every collector at length awakens from 
his self-induced trance to admit a need for 
setting a limit to his enterprise. So, after 
Philadelphia, I decided to use the two long 
lists only once more, in London. But I 
couldn't drop certain unfound probabilities. 
Therefore I made a list of about 50 choice 
ones and set out to find them in the direc- 
tories for Baltimore, Bronx, Cleveland, 
Houston, Long Beach, Manhattan, Milwau- 
kee, Minneapolis, NeAvark, New Orleans, 
Portland (Oregon), Queens, St. Louis, San 
Diego, Seattle, and Washington. There 
were satisfactions in this search, since most 
of the surnames that turned up may be 
considered to be rarities. First, the plants, 
the last one being more recent than most 
biological surnames: 

Bellrose Scallion Tobacco 


There was one fish, Girling (a young sal- 
mon), as well as several other vertebrates: 
Chicken Mouse Pups 

And some interesting anatomy: 

Fishes were represented by : 






Examination of the London Telephone 
Directory, of 2,209 three-column pages, 
from Aaggard to Zyrot, followed. I was in 
effect searching the "mother lode," the orig- 
inal source from which family names of 
English-speaking people have been scat- 
tered over the world. None of the following 
had been found in American lists. 

Botanical surnames were : 










And, as always, a few additions to the 
long list of fishes: 

Gargett Ilaice Smelt 

Grayling Pout Sucker 

One amphibian, Newte, and three birds: 
Condor Gooney Phoebe 

The list of mammals included several of 
the few names of domesticated animals that 
had not been located before: 












Anatomical terms were Beak, Chine, 
Hide, Skinn, Skull, Toe, and Wrist. 

Invertebrates found in all directories are 
combined in a single list, which includes a 
few surnames associated with parts or prod- 
ucts of backboneless animals. 








































In the course of the investigation I re- 
corded a number of surnames which have 
their origin in mythological persons and 















Not so much mythological as fictional, 
Doe (John) and Roe (Richard), long prom- 
inent in the symbolism of the law, both 
refer to names of deer. Other imaginary 
persons were Punch and Judy. 

Even after the good returns from my 
hunt through the London books, I was not 
satisfied. Missing were many surnames that 
could have escaped transfer from the names 
of common living things only by some im- 
probable accident. I had located more than 
300 fishes, amphibians, reptiles, birds, and 
mammals, but where were Donkey, Heifer, 
Hinny, Horse, Sheep, Toad, and Woman? 
About 100 family names were based on 
animal structures, but why had Jowl, 
Stomach, and Tail eluded search? Plant 
surnames to the number of about 200, not 
to mention some 35 terms for plant parts, 
had turned up, but not Lentil and Turnip. 
A fair representation of the small but useful 
invertebrates did not include the super- 
latively toothsome Lobster. 

Are some of the "lost" families still occu- 
pying the same English farms where their 
family trees became established hundreds 
of years ago? Perhaps certain of the male 
lines have run out, or other names have 
been chosen. Possibly some are living in the 
United States outside of large cities. They 
may have no telephones. Some of them may 
be living in other countries. 

According to an analysis of surnames re- 
corded during the first United States Cen- 
sus, 2 the following families which I have 
been unable to find in telephone books were 
living in this country in 1790: 

Bantam Horse Sealion 


It is of interest to record that surnames 
found in the course of the First Census 
were 83.5 percent English and 91.8 percent 
from the British Isles. 

In a final effort to find Horse, Sheep, and 
Turnip I looked in the telephone direc- 
tories of Atlanta,, Buffalo, Cincinnati, 
Columbus, Dallas, Dayton, Denver, Fort 
Wayne, Fort Worth, Indianapolis, and 
Pittsburgh; Montreal, Toronto, and Van- 

2 A century of population growth. From the First 
Census of the United States to the Twelfth, 1790- 
1900. U. S. Department of Commerce and Labor, 
Bureau of the Census. 303 pp., 1909. 

Mar. 15, 1948 little: new species of trees from western Ecuador 


couver, Canada; Melbourne and Sidney, 
Australia. Although the Turnipseed family 
is well distributed, Turnip still eluded me. I 
could get no closer to Horse than Horsey, 
which occurs plentifully. But I found Sheep. 
That is my vara avis; or, more accurately, 
vara ovist 

Those who write reports of the results of 

their research seldom mention one influence 
that helps to shape the course of their en- 
deavors. I refer to the comments of wives. 
In my case attention was called to the 
possibility of sometime bringing the search 
to a close by a gentle remark, "I'm getting a 
little tired of seeing piles of telephone books 
around the house." 

BOTANY. — New species of trees from western Ecuador. 1 
Forest Service, U. S. Department of Agriculture. 

Elbert L. Little, Jr., 

The 19 new species here described and il- 
lustrated are from the author's collection of 
tree specimens made in 1943 in western 
Ecuador with the Latin American Forest 
Resources Project of the U. S. Forest Serv- 
ice. Genera represented are: Heisteria, Hir- 
tella (2), Brownea (3), Humiria, Erythro- 
chiton, Tetragastris, Guarea, Trichilia, Be- 
lotia, Matisia (2), Clusia (2), Neosprucea, 
Symplocos, and Aspidosperma. Twelve spe- 
cies are from the province of Esmeraldas, 
and others from the provinces of Pichincha, 
Chimborazo, Los Rios, and El Oro. Addi- 
tional new species from this collection are 
being published elsewhere by specialists, 
and others must remain undescribed at 
present because of incomplete material. 

Type specimens have been deposited in 
the U. S. National Herbarium, Washington, 
D. C, and a set of isotypes in the herbarium 
of the U. S. Forest Service, Washington, 
D. C. Additional isotypes of most species 
are in the folloAving herbaria: Tropical 
Forest Experiment Station, U. S. Forest 
Service, Rio Piedras, Puerto Rico; Yale 
University School of Forestry, New Haven, 
Conn.; Chicago Natural History Museum, 
Chicago, 111.; and Instituto Botanica, Uni- 
versidad Central, Quito, Ecuador. 

Acknowledgment is due C. V. Morton, 
of the U. S. National Museum, for assist- 
ance in checking the Latin diagnoses, to 
Ellsworth P. Killip, also of the National 
Museum, for suggestions, to Miss Leta 
Hughey, botanical artist, Forest Service, 
for making the drawings of the new species, 
and the late Dean Samuel J. Record, Yale 
University School of Forestry, for making 
generic determinations of wood samples of 

1 Received October 3, 1947. 

four of these new species. The Forest Serv- 
ice project in Ecuador is described in the 
following report: Holdridge, L. R., et al., 
The forests of western and central Ecuador. 
134 pp., illus. Forest Service, U. S. Dept. 
Agr., Washington, D. C. 1947. (Processed.) 

Heisteria cyathiformis Little, sp. nov. 
Fig. 7 

Sec. Euheisteria Engl. Frutex vel arbor parva 

4 m alta, trunco 3 cm diametro, glabra, ramulis 
gracilibus teretibus viridulis, novellis subangu- 
latis, internodiis brevibus 0.5-3 cm longis. Folia 
petiolata, petiolis 4-11 mm longis, supra leviter 
canaliculars; laminae anguste ellipticae vel 
oblanceolatae, (5) 8-15 cm longae, 3-5.5 cm 
latae, basi acutae vel acuminatae, apice sub- 
abrupte acuminatae, acumine 4-9 mm longo, 
membranaceae, margine integrae, costa supra 
plana subtus elevata, nervis lateralibus utrin- 
que 5-9, prope marginem arcuato-conjunctis, 
subtus prominulis. Flores fasciculares in axillis 
2 vel 3; pedicellus 3 mm longus, apicem versus 
crassus; albastrum ca. 1.5 mm longum; calyx 
1 mm longus ad medium 5-lobus, lobis triangu- 
laribus acutis; corolla in alabastro globoso- 
ellipsoidalis, 1.3 mm longa, viridulo-albida, 5- 
lobata, intus minute puberulenta; stamina 10, 

5 longiora et 5 alternatim breviora ; ovarium tri- 
angulare, apicem versus acuminatum, 1 mm 
longum, 1 mm diametro, 3-loculare ovulis 3, 
stigmate minuto sessili 3-lobato. Fructus soli- 
tarius, pedicello in stato fructifero 12-15 mm 
longo, calyce fructifero accrescente cyathi- 
formi, 4-5 mm longo et 7 mm lato, truncato, 
subintegro vel minute 5-dentato, rubro; drupa 
ellipsoidalis, 11 mm long, 7 mm diametro, 

Shrub or small tree 4 m tall and 3 cm in 
trunk diameter, said to become larger. Speci- 



mens collected from two plants in wet tropical 
forest at Quininde\ "Aji." 

Esmeraldas: Quininde, altitude 65 in, April 
15, 1943, Little 6263A (Forest Service no. 
96832; U. S. Nat. Herb. no. 1854915, type). 
Little 6263 (Forest Service no. 96833). 

Heisteria Jacq. has more than 40 species in 
South and Central America and 3 in western 
Africa. Most species have the enlarged calyx 
of the fruit lobed and rotate, reflexed, or loosely 
enclosed the drupe. In H. cyathijormis the red, 
cup-shaped or hemispherical calyx is about one- 
third the length of the black, ellipsoidal drupe 
(red when immature) and closely adheres to 
the lower part of the drupe. This enlarged 
fruiting calyx is truncate and subentire or 
minutely 5-toothed. Superficially the fruit, 
with its elongated drupe and cuplike calyx, 
resembles that of some members of the Laura- 
ceae, though the other characters are unlike 
that family. Caroline K. Allen and I. W. Bailey 
kindl} r have suggested that this material prob- 
ably belongs in Heisteria, where it apparently is 

Heisteria cyathijormis is related to H. cyano- 
carpa Poepp. & EndL, of the Amazonian region 
of Brazil, Peru, and Ecuador, which has larger 
and thicker leaves, more flowers in a fascicle, 
and the fruiting calyx broad, orbicular, usually 
reflexed. The collection contains material of 
another species of this genus, doubtfully re- 
ferred to the Central American species H. mac- 
rophylla Oerst. 

Hirtella carbonaria Little, 
Pirf. 3 

sp. nov. 

Arbor medio cris 12 m alta, trunco 25 cm 
diametro. Cortex laevis, griseus. Ramuli parce 
hirsutuli mox glabrescentes. Folia petiolata 
cum stipulis linearibus, ca. 3 mm longis, in- 
volutis, strigosis; petiolis ca. 2 mm longis, parce 
hirsutulis; laminae ellipticae, 5-9 cm longae et 
2.5-4 cm latae, basi rotundatae, apice abrupte 
acuminatae, integrae, inflexae, coriaceae, re- 
ticulatae, supra glabrae, subtus in costa et 
venis pilis paucis subappressis praeditae, venis 
primariis 5-8-jugis, arcuatis et margine anasta- 
mosantibus, his cum costa utrinque paullo ele- 
vatis. Inflorescentiae terminales atque axillares, 
thjrrsiformes, multiflorae, cum floribus multis 
abortivis, 6-9 cm longae et 2-4 cm latae, pe- 
dunculo 2-5 cm longo, rhachibus pilos paucos 
appressos gerentibus; bracteae triangulares, 

acutae, 1-2 mm. longae; pedicelli 3 mm longi; 
hypanthium campanulatum, 3 mm longum et 
2 mm latum, externe sparse pubescens, intus 
glabrum; sepala 5, elliptica, obtusa, 3 mm longa 
et 2.5 mm lata, externe strigillosa, intus densius 
strigillosa; petala 5, alba, elliptica, 4 mm longa 
et 2.5 mm lata, retusa, glabra; stamina 3, 8- 
12 mm longa, filamentis glabris, basi connatis; 
staminodia 2, 1.5 mm longa; ovarium fere 2 
mm longum, dense pilosum; stylus deorsum 
sparse pilosus, 10-15 mm longus. Drupa ob- 
ovoidea, 15 mm longa, 8 mm lata, 3.5 mm 
crassa, compressa, atra, glabrescens, intus 
dense pilosa; semen ellipsoideum, compres- 
sum, fuscum, ca. 10 mm longum, 5 mm latum, 
2.5 mm crassum. 

Medium-sized tree to 12 m tall and 25 cm 
in trunk diameter. Bark smooth, gray. Wood 
used for charcoal, as the common and specific 
names indicate. Common in moist soil, flood 
plain forest near Rio Nadadero, wet tropical 
forest zone at San Lorenzo. "Carbonero," 

Esmeraldas: San Lorenzo, altitude near 
tidewater, April 19, 1943, Little 6275 (Forest 
Service no. 96790; U. S. Nat. Herb. no. 
1858838, type). 

Besides Hirtella triandra Sw., this collection 
contains two undescribed species of Hirtella L. 
Hirtella carbonaria has elliptic, coriaceous, 
glabrescent leaves that are turned inward 
slightly at the midrib. The pubescence of twigs, 
leaves, and inflorescence is rather sparse, con- 
sisting generally of a few appressed hairs, which 
are lost as the parts become nearly glabrous 
with age. The thyrsiform inflorescence contains 
many abortive flowers. 

Hirtella pauciflora Little, sp. nov. 
Fig. 2 

Arbor magna 25 m alta, trunco 50 cm diame- 
tro. Cortex laevis, leviter fissilis, fuscus. Ramuli 
dense hirtelli. Folia petiolata cum stipulis 
linearibus, ca. 3 mm longis, dense hirtellis; 
petiolis 3-4 mm longis, dense hirtellis; laminae 
ellipticae, 7-13 cm longae et 3-5 cm latae, 
basi acutae, apice acuminatae, integrae, mem- 
branaceae, pallide virides, supra costis his- 
pidae, pilis paucis rigidis in venis et venulis 
ferentes, mox glabratae, subtus in venis et 
venulis hispidae, nervis primariis 6-9-jugis, 
ascendentibus, his cum costa supra leviter im- 
pressis, subtus prominentibus. Inflorescentiae 

Mae. 15, 1948 little: new species of trees from western Ecuador 


Fig. 1. — Belotia australis, f X; flower, 2X; fruits, f X. Fig. 2. — Hirtella pauciflora, f X; flower, 

2 X . Fig. 3. — Hirtella carbonaria, f X ; flower, about 3 X ; fruit, -J X . Fig. 4. — Guarea polymera, 

|X; flower, IX; fruit, IX. 



racemosae terminates in ranmlis brevibus, in- 
conspicuae, 2- vel 3-florae, 3-6 cm longae, 
pedunculo hirtello, 1-4 cm longo; pedicelli 
crassi, inflati, lignei, 10-12 mm longi, basi fere 

1 mm diametro, medio et apice 1.5-2 mm di- 
ametro, dense hirtelli; flores viridulo-fusci, ca. 
5 mm longi et 5 mm diametro; hypanthium en. 

2 mm longum et 2 mm latum, externe hir- 
tellum, intus fauce dense retrorse pilosum, alibi 
glabrum; sepala 5, late elliptica, obtusa, 3 mm 
longa et 2-3 mm lata, externe dense hirtella, 
intus dense puberula; petala 5, elliptica, obtusa 
vel retusa, 3 mm longa et fere 2 mm lata; 
stamina 3, filamentis glabris, basi connatis, 
5-7 mm longis; staminodia 2, minuta, 0.2 mm 
longa; ovarium fere 2 mm longum, dense pi- 
losum; stylus basin versus sparse pilosus, alibi 
glaber, 4 mm longus. Fructus deest. 

Large tree 25 m tall and 50 cm in trunk di- 
ameter. Bark smooth, slightly cracked, brown. 
Virgin wet tropical forest and old cacao planta- 
tion at Pichilingue. "Coquito." 

Los Rfos: Pichilingue, altitude 45 m, May 
21, 1943, Little 6475 (Forest Service no. 95855; 
U. S. Nat. Herb. No. 1876218, type). 

The inconspicuous, small racemes terminal 
on short branches and bearing only 2 or 3 
flowers distinguish H. pauciflora. 

Brownea disepala Little, sp. nov. 
Fig. 9 

Arbor magna 30 m alta; trunco 35 cm di- 
ametro, florifero, cum tumoribus ellipsoidalibus 
magnis et conspicuis, 3-5 cm crassis, 3-5 cm 
longis, et 3-10 cm latis, cicatricibus inflores- 
centiarum ferentibus. Cortex griseus, fere 
laevis; ramuli glabri. Folia paripinnata; petioli 
crassi, subglabrati, 6-15 mm longi et 3-4 mm 
diametro; rhaches 20-40 cm longi, fere glabri; 
petioluli crassi, corrugati, 3-5 mm longi, sub- 
glabri; foliolae (4) 6-9-jugae, oppositae, ob- 
longo-ellipticae, (5) 9-12 cm longae, (2) 3-4 
cm latae, basi inaequales, apice longe caudato- 
acuminatae, subcoriaceae, nitidae, reticulatae, 
glabrae, costae basi glande uno minuto 
gerentes. Inflorescentia caulina capitata, stro- 
biliformis, globosa, sessilis, pallide incarnata, 
5 cm diametro; bracteae multae, subrotundae 
vel ovatae vel oblongae, valde concavae, 1-7 
cm longae, externe dense tomentosae, intus 
glabrae vel leviter pilosae, exteriores satis 
crassae et coriaceae, interiores membranaceae; 
rhachis 2.5-3 cm longa et 8-10 mm diametro, ad 

f longitudinis bracteata, apieem versus flori- 
fera. Flores multi (ca. 30-40), albidi; pedicelli 
10-18 nun longi, pilosuli; vagina nulla; hy- 
panthium campanulatum, subtetragonum, 8 

mm longum, 5-6 mm latum, pilosulum; sepala 
2, oblanceolata vel oblonga, apice bidentata 
dentibus 3-5 mm longis, pilosula vel glabrata, 
nervis parallelis, 30-50 mm longa et 10-15 
mm lata; petala 4, longe angusteque unguicu- 
lata, glabrata, pinninervia, 2 majora oblance- 
olata, obtusa, 37-43 mm longa et 10 mm lata, 
2 minora lineari-spathulata, acuta, 25-27 mm 
longa et 3 mm lata; vagina staminalis 3 cm 
longa, pilosula; stamina 17 vel 18, filamentis 
25-50 mm longis, antheris 5-6 mm longis; 
ovarium stipitatum (6 mm), 15 mm longum, 
2.5 mm latum, compressum, dense tomento- 
sum; stylus glaber, 50-60 mm longus; stigma 
capitellatum. Legumen paucum, ca. 15 cm 
longum, 4 cm latum, plano-compressum, rufo- 
puberulum, valvis post dehiscentiam tortuosis. 

Large tree 30 m tall and 35 cm in trunk diam- 
eter, distinguished by cauliflory. Bark gray, 
smoothish. Trunk easily recognized by the en- 
larged elliptical swellings at scars of former in- 
florescences. The large whitish flowers are borne 
in a headlike cluster from a spherical pink, 
bracteate, strobiliform bud about 5 cm in di- 
ameter which is sessile on the trunk. From the 
flower cluster a few large flat pods develop. 
Mountain (wet tropical) forest in southwestern 
Ecuador. "Palo de vaina." 

El Oro : Near Piedras, altitude about 800 m, 
June 21, 1943, Little 6677 (Forest Service no. 
98587; U. S. Nat. Herb. no. 1857106 and 
1857107, type). Guayas: Near Pedro Carbo, 
altitude about 300 m, Jan. 13, 1943, Oscar 
Haught 3131 (U. S. Nat. Herb.). 

H aught 3131 has flowers similar to those of 
the type but differs in having the leaflets 
slightly smaller (5-7 cm long and 2-2.5 cm 
broad). It is described by the collector as a 
moderate-sized tree, strictly cauliflorous, with 
white flowers, in a forest on calcareous soil. 

The genus Brownea Jacq. contains more than 
25 known species, principally from Venezuela, 
Colombia, and Peru, with more from Venezu- 
ela than any other country. Some species occur 
also in Panama, the Guianas, and Brazil. 
Many of these are of restricted distribution. 
Only one new species, B. herthae Harms, has 
heretofore been described from Ecuador. In 
addition to that species, this collection contains 

Mar. 15, 1948 little: new species of trees from western Ecuador 


Fig. 5. — Brownea angusti flora, iX; flowers, about f X. Fig. 6. — Tetragastris varians. \X; flower, 

4X. Fig. 7. — Heisteria cyathiformis, |X. Fig. 8. — Brownea puberula, |X. 



three new species of Brownea. Most species of 
Brownea have very showy, brilliant red flowers 
borne in terminal inflorescences. Very few spe- 
cies have white flowers and very few have cauli- 
flory. Brownea disepala is odd in both ways in 
having white flowers borne in clusters on the 

The segregate genus Browneopsis Huber 
(Bol. Mus. Goeldi 4: 565. 1906) was proposed 
for the species with sheath absent, larger num- 
ber of stamens, and lesser development of 
petals (lacking, rudimentary, or 4). However, 
Ducke, Record and Hess, and Macbride have 
reduced the segregate to synonymy. Besides 
Brownea disepala, other species in this group 
are: Brownea ucayalina (Huber) Ducke, Brown- 
ea cauliflora Poepp. & EndL, Brownea excclsa 
(Pittier) Macbr., and Brownea peruviana 
Macbr. Of these, the new species is like the 
last in its calyx. Macbride (Field Mus. Bot. 13, 
pt. 3, no. 1: 134-135, 132. 1943) in the original 
description noted that B. peruviana was aber- 
rant in having 4 petals and 2 colored, petaloid 
bractlets or sepals. After examining a specimen 
of one of the two collections cited by Macbride 
(Klug 3810), I interpret as sepals the two paral- 
lel nerved structures of both species attached in 
the usual position of the calyx, at the summit of 
the hypanthium and outside the four clawed, 
pinnately nerved petals. These sepals, which 
in both species are slightly 2-toothed or 2-lobed 
at the apex, may represent a reduction from 4 
sepals to 2 connate ones. This interpretation is 
simpler than the alternate one also mentioned 
by Macbride of regarding the petals as absent, 
the 4 clawed petaloid structures as bractlets. 
In Brownea the bractlets or sheath, when 
present, are attached at the base of the hypan- 
thium, not at the summit. 

Brownea puberula Little, sp. nov. 
Fig. 8 

Arbor 6 m alta vel major, trunco 10 cm di- 
ametro. Ramuli, petioli, rhaches, et petioluli 
dense et minute fusco-puberuli vel demum 
glabra ti. Folia paripinnata; petioli basi crassi, 
5-40 mm longi; rhaches (3) 7-25 cm longae; 
petioluli 3-7 mm longi; foliolae (1) 3-7-jugae, 
oppositae, oblongo-ellipticae vel ovatae, (4) 
7-12 cm longae, (2) 3-6 cm latae, basi inae- 
quales, obtusae vel rotundatae, apice longe 
caudato-acuminatae (1.5-2.5 cm longa), cori- 

aceae, supra glabrae, subtus minute puberulae, 
reticulatae, costae basi glande uno gerentes. 
Inflorescentiae terminales vel laterales, inter- 
dum ex rhachibus inflorescentiarum prioriuin; 
racenii breves, umbelliformes, rhache 2 cm 
longa, bracteis caducis; flores 15-25, rubri vel 
coccinei, spectabiles; pedicelli tenues, puberuli, 
10-20 mm longi; vagina petaloidea tubulosa, 
bilabiata, 30-32 mm longa, externe appresso- 
puberula, intus glabra, lobis acutis 10-12 mm 
longis; hypanthium anguste tubulosum, 12-14 
mm longum, 3 mm diametro, puberulentum vel 
glabratum; sepala 4, petaloidea, glabra, 24-27 
mm longa, anterius obovatum, acutum, 8-11 
mm latum, alia oblanceolata, obtusa, 4-6 mm 
lata; petala 5, fere aequalia, 37-46 mm longa, 
9-13 mm lata, longe angusteque unguiculata, 
glabra, obovata, obtusa; stamina 11, 38-45 
mm longa, ad \ vel f longitudinis connata; 
vagina staminalis externe glabra, intus basi 
pilosa; filamentis glabris, 10-18 mm longis; 
antheris 2 mm longis; ovarium 7-8 mm long- 
um, dense tomentosum; stylus tenuis, deorsum 
tomentosus, 27-35 mm longus; stigma capitel- 
latum. Legumen 20 cm longum, 4.5 cm latum, 
planocompressum, minute fusco-tomentosum, 
valvis post dehiscentiam tortuosis. 

Tree 6 m tall and 10 cm in trunk diameter. 
Uncommon in understory of wet, swampy 
tropical forests at sea level, including flood 
plain at edge of river and higher level of man- 
grove swamp forest at San Lorenzo. 

Esmeraldas: San Lorenzo, sea level, April 
22, 1943, Little 6338 (Forest Service no. 98230; 
U. S. Nat. Herb. no. 1909574, type); April 22, 
1943, Little 6338 A (Forest Service no. 98231); 
April 19, 1943, Little 6291 (Forest Service no. 
96719). Province not known: Moist forest at 
foot of Western Cordillera, altitude 200 m, 
Rimbach 183 (U. S. Nat. Herb.). 

Rimbach 183, from an inland locality, was 
described by the collector as a tree 20 m high 
and 25 cm in trunk diameter, with red flowers 
in dense umbels, mostly from old wood, some- 
times near base of trunk, and more rarely from 
the branches. 

This species is unusual in the dense, minute, 
dark brown pubescence of the young branches, 
petioles, rachises, and petiolules. However, in 
age the pubescence may be nearly absent. The 
inforescences are fewer flowered and the flowers 
more nearly glabrous than in most species of 
Brownea. Though it has the bilabiate floral 

Mar. 15, 1948 little: new species of trees from western Ecuador 


sheath found in most species of the genus, 
Brownea puberula is distinct in the size, shape, 
and pubsecence of the floral parts. It resembles 
B. birschellii Hook, f., of Venezuela, which is 
glabrous and has much larger clusters of larger 
flowers. In publishing that species, J. D. 
Hooker (Bot. Mag. 98: pi. 5998. 1872) men- 
tioned that "it closely resembles a species from 
the Pacific sea-coast of New Granada, at Es- 
meraldas, collected by Colonel Hall, and pre- 
served in the Hookerian Herbarium." Hall's 
specimen, collected near the type locality of 
B. puberula, upon examination may prove to 
be a much earlier collection of this species. 

Brownea angustiflora Little, sp. no v. 
Fig. 5 

Arbor parva 10 m alta, trunco 15 cm diame- 
tro. Ramuli glabri. Folia paripinnata, glabra; 
petioli basi crassi, 4-25 mm longi; rhaches 3-15 
cm longae; petioluli 3-5 mm longi; foliolae 2- 
vel 3-jugae, oppositae, oblongo-ellipticae, (7) 
10-23 cm longae, (3) 4-7 cm latae, basi inae- 
quales, obtusae velacutae, apice caudatoacumi- 
natae (1-3 cm longo), coriaceae, reticulatae, 
costae basi glande uno gerentes. Inflorescentiae 
terminales et laterales vel caulines; racemi 
breves, umbelliformes, rhache 1.5-2 cm longa, 
bracteis caducis; flores ca. 15-20, rubrae vel 
coccineae, spectabiles; pedicelli tenues, pu- 
beruli, 9-12 mm longi; vagina petaloidea, 
anguste tubulosa, 35-52 mm longa, 6 mm 
diametro, uno latere ad 15-20 mm fissa, externe 
puberula, intus glabra; hypanthium anguste 
tubulosum, 11-15 mm longum, 2-2.5 mm 
diametro, glabrum; sepala 4, petaloidea, glabra, 
anterius oblanceolatum, acutum, 35 mm lon- 
gum et 6-7 mm latum, alia linearia, acuta, 28- 
30 mm longa et 2-3 mm lata; petala 5, longe 
angusteque unguiculata, glabra, maximum ob- 
lanceolatum, retusum vel apiculatum, 42 mm 
longum, 6-8 mm latum, duo oblanceolata, 
retusa, 35-38 mm longa, 4-6 mm lata, duo 
linearia, obtusa, 20-22 mm longa, 1-2 mm lata; 
vagina staminalis 25-29 mm longa, glabra vel 
leviter pilosa; stamina 9 vel 10, filamentis 10- 
20 mm longis, antheris 2 mm longis; ovarium 
9-10 mm longum, 2 mm latum, dense tomento- 
sum; stylus tenuis, inferne tomentosus, 35-40 
mm longus; stigma capitellatum. Legumen 
paucum, 20 cm longum, 4 cm latum, plano- 
compressum, minute puberulum; valvis post 

dehiscentiam tortuosis; semina orbiculata, 
plana, 2.5 cm diametro. 

Small tree 10 m tall and 15 cm in trunk 
diameter, in understory of wet tropical forest 
and in clearing. "Flor de mayo." 

Esmeraldas: Borbon, elevation 15 m, May 
3, 1943, Little 6418 (Forest Service no. 98232; 
U. S. Nat. Herb. no. 1879851, type); April 26, 
1943, Little 6382 (Forest Service no. 98225). 

Two collections of this species were made, 
both at the same locality. In the type collec- 
tion, Little 6 41 8, the inflorescences are terminal 
or lateral on the leafy branches. Little 6382 
has the flowers borne on the trunk and slightly 
larger and more mature. 

This species is characterized by long, narrow, 
slender, nearly glabrous flowers in few flowered 
clusters. The narrowly tubular flower sheath, 
which is cleft on one side, is unusually long, 35- 
52 mm long, and nearly encloses the perianth, 
though the three larger petals are exserted. The 
sheath generally varies from 12 to 32 mm long 
in other species of Brownea. Only a few others, 
such as B. coccinea Jacq. and B. aroensis Pit- 
tier, have the sheath cleft on only one side 
instead of bilabiate. 

Humiria procera Little, sp. nov. 
Fig. 12 

Arbor magna recta, 30-34 m alta vel major. 
Ramuli glabri, 4-alati ex foliis et costis decur- 
rentibus, alis 1-2 mm latis. Folia disticha, de- 
currentia, sessilia, coriacea, vivo atroviridia, 
nitida, late elliptica, 8-13 cm longa et 5-7 cm 
lata, basi obtusa, apice abrupte et breviter 
acuminata (5-10 mm), margine remote glandu- 
loso-crenulata, supra glabra, subtus minute 
rugulosa, costa subtus carinata, carina angusta, 
basi 1 mm alta, decurrente. Inflorescentia axil- 
laris, corymbosa multiflora, 4-6 cm longa et 
3-5 cm lata, puberula, ca. 5 vel 6 dichotoma 
vel trichotoma; bracteae triangulares, acutae, 
1-2 mm longae; pedunculus alatus anceps, 2- 
2.5 cm longus; pedicelli 1 mm longi; flores 
minuti, viriduli, ca. 2 mm longi et 1.5 mm lati; 
sepala 5, rotundata, ciliata, leviter puberula, 
1 mm longa; petala 5, oblonga, externe leviter 
puberula, 1.5-2 mm longa et 0.5 mm. lata; 
stamina 20, glabra, ca. 1.5 mm longa; fila- 
menta basi connata; cupula hypogyna e 
squamis plus minus connatis composita; pistil- 
lum ca. 1.5 mm longum; ovarium ovoideum, 
parce puberulum, 5-loculare, ovulis 2 in loculo; 



stylus crassus; stigma 5-radiatum. Drupa 
ovoidea vel cllipsoidea, base obtusa, apice 
acuta vel obtusa, tomentosa, 25-35 mm longa, 
17-20 mm lata; endocarpium osseum, sub 
apice 5-foveolatum. 

Large erect tree of forest canopy in wet 
tropical forest, 30 to 34 m tall and 40 cm in 
diameter, according to measurements from 
trees felled for wood samples, but becoming 
larger. The fruits resemble pecans (Carya il- 
linoensis (Wangenh.) K. Koch) in size and 
shape. It is said that the oily seeds are edible. 

Esmeraldas: Playa de Oro, altitude about 
65 m, May 1, 1943, Little 6^12 (Forest Service 
no. 95906; U. S. Nat. Herb. no. 1876214, type; 
with wood sample). Little 64-13 (Forest Service 
no. 95907; with wood sample). San Lorenzo, 
altitude about 10 m, April 21, 1943, Little 
6320 (Forest Service no. 96880; with wood 
sample). Quininde, altitude 65 m, April 9, 
1943, Little 6233 (old fruits only). 

Common and probably widely distributed in 
the wet tropical forest of Esmeraldas. In some 
locations near San Lorenzo it comprises a large 
part of the big timber. Very common also along 
the banks of Rio Santiago above Selva Alegre, 
where Chanuzal, a group of houses, apparently 
derives its name from the man}'' trees of 
"chanul." After partial cutting of other tree 
species, "chanul" occurs in nearly pure stands. 
It exists in quantities along the main rivers and 

The South American genus Humiria Jaume 
St.-Hil. is distributed in Brazil, the Guianas, 
Venezuela, and Colombia and apparently has 
not been reported previously from Ecuador. 
Though the characteristic size is that of shrubs 
and small trees, two species are described as 
becoming large trees under favorable condi- 
tions. Selling (Svensk Bot. Tidskr. 39: 257-269, 
illus. 1945) distinguished 6 living species and 3 
fossil ones. The latter, based upon fruits from 
Brazil, Colombia, and Peru (Piura), indicated 
a broader distribution of the genus in the Ter- 
tiary than at present. 

Wood samples of the Ecuadorian material 
were first determined by S. J. Record as 
Humiria sp. near H. floribunda Mart., which 
occurs in the Guianas, Amazonian Brazil, 
Venezuela, and Colombia. Subsequent study 
of herbarium specimens confirmed this deter- 
mination and showed that an undescribed spe- 

cies was represented. Humiria procera is readily 
separated from other species of the genus by 
the numerous minute flowers only about 2 mm 
long, while the others have fewer, larger flowers 
usually about 4-5 mm long. In H. procera 
the twigs are prominently 4-winged and the 
leaves are distichous, abruptly acuminate, and 
crenulate, while in H. floribunda the twigs are 
only slightly winged, the leaves spirally ar- 
ranged, smaller, usually obtuse at apex, and 
entire or nearly so. The distichous leaves dis- 
tinguish H. procera from most species of the 
genus also. 

The following notes on the wood of Humiria 
procera were made by J. Edson Myer, of the 
Forest Service field party. Sapwood grayish 
white, about 5 cm thick. Heartwood dull red, 
becoming brownish on exposure, hard and 
heavy, having a specific gravity of 0.68 based 
upon volume when green or 0.82 based upon 
volume when oven dry. Texture fine and uni- 
form. Growth rings are indistinct due to fewer 
pores in denser zones; the pores are minute, 
thin walled, and numerous. The rays are very 
fine, not visible on cross section and low and 
inconspicuous on radial section. Wood fairly 
difficult to cut, rather harsh, readily polished, 
strong but brittle, and of medium durability. 
Not used much because of alleged silica content 
which dulls saws. Might be used for construc- 
tion where not exposed to the weather. 

Waterman (Trop. Woods 88: 1-11. 1946 
(1947)) included wood samples of the collec- 
tions cited above in tests of decay resistance of 
water-soluble extractives of several tropical 
American woods to growth of wood-decay 
fungi. This species, listed as Humiria sp., was 
found to be nontoxic or not resistant to decay. 

Erythrochiton carinatus Little, sp. no v. 
Fig. 14 

Sect. Toxosiphon (Baill.) Engl. Frutex vel 
arbor parva 2-5 m alta, trunco 5 cm diametro. 
Cortex griseus, leviter asper et fissilis, decorti- 
cans. Folia solitaria, unifoliata; petioli (2) 
4.5-7 cm longi, 1.5 mm diametro, glabri, 
tenues, semiteretes, apice tumido-articulati; 
laminae oblongo-ellipticae, 11-20 cm longae et 
5-8 cm latae, basi cuneatae, apice abrupte 
acuminatae, integrae, papyraceae, glabrae, in- 
conspicue glanduloso-punctatae, venis lateral- 
ibus primariis ca. 12-14, angulo obtuso egredi- 
entibus, arcuatis, venulis reticulatis. Inflores- 

Mae. 15, 1948 little: new species of trees from western Ecuador 


Fig. 9. — Brownea disepala, £ X ; axis of old inflorescence, | X ; young globose, unopened inflorescence, 
|X; flower, \ X. Fig. 10. — Trichilia floribunda, £ X; bud, 6X; flower, 6X. 



centia terminalis, 17-19 cm longa, longepedun- 
culata; pedunculus 11-15 cm longus, sulcatus, 
strigillosus, apice furcatus, ca. 3-florus; flores 
desunt. Sepala libera, persistentia, bractei- 
formia, viridia, lanceolata, acuminata, reticu- 
lata, inconspicue glanduloso-punctata, externe 
fere glabra, intus dense tomentulosa, in fructu 
30-37 mm longa et 4-8 mm lata, capsulis sub- 
duplo longiora; capsula fere ad medium 5- 
lobata, pallide viridis, 15 mm alta et ca. 20 mm 
lata, coccis 5, reticulatis, inconspicue glandulo- 
so-punctatis, puberulis, ca. 15 mm altis et 9 
mm latis, carina prominente 1-2 mm lata, 
mucrone 1 mm longo; semina 2, superposita, 
fusca, fere laeve, 5 mm longa et -4 mm lata et 

Shrub or small tree 2 to 5 m tall and 5 cm 
in diameter. Bark gray, slightly rough and fis- 
sured, shredding. Common locally in under- 
story of mountain (wet tropical) forest. 

El Oro: Piedras, altitude about 500 m, 
June 18, 1943, Little 6632 (Forest Service no. 
98509; U. S. Nat. Herb. no. 1909573, type). 

The genus Erythro chiton Nees & Mart, con- 
tains seven previously described species of 
shrubs or small trees widely but sparingly dis- 
tributed in tropical America from southern 
Mexico to Colombia, Brazil, and Peru. Engler 
(in Engler and Prantl, Natiirl. Pflanzenfam. ed. 
2, 19a: 288. 1931) listed five species, and two 
others have been added. Apparently this genus 
has not heretofore been reported from Ecua- 
dor. Specimens of both Erythrochiton brasilien- 
sis Nees & Mart, and this new species were 
collected near Piedras, El Oro. 

Erythrochiton carinatus belongs to Section 
Toxosiphon (Baill.) Engl., in which the 5 
sepals are free, and is distinguished in fruit 
from the other four species of this section by the 
elongate, acuminate sepals about twice the 
length of the capsule and by the deeply lobed 
capsule, with a prominent keel on each seg- 

Tetragastris varians Little, sp. nov. 
Fig. 6 

Arbor 25 m alta, trunco 50 cm diametro. 
Ramuli crassi, subteretes, brunnescentes, glab- 
ri, lenticellis elevatis. Folia alterna, pinnata, 
apicem ramulorum versus trifoliolata vel uni- 
foliolata, petioli 4-9 (21) cm. longi, evidenter 

sulcati, subancipites, glabri, supra plani; 
rhaches foliorum trifoliolatorum 2-4 cm longae, 
sulcatae, glabrae, supra planae; petioluli 2-4 
mm longi; laminae foliolarum late ellipticae ad 
obovatae, 9-20 cm longae et 5-10 cm latae, 
basi obtusae vel cuneatae, apice abrupte 
acuminatae (5-10 mm), chartaceae, integrae, 
venis primariis utrinque latere ca. 9-12, angulo 
lato divergentibus, arcuatis et prope marginem 
anastomosantibus, utrinque venis et venulis 
elevatis et prominente reticulatis, supra 
glabrae, subtus glabratae, minute papillosae, 
ad costam pilis paucissimis appressis praeditae. 
Paniculae laterales apicem ramulorum versus, 
7-12 cm longae, multiflorae; rhaches leviter 
compressae, minute strigillosae, bracteatae; 
bracteae triangulares, acutae, 1 mm longae; 
pedicelli 0.5-1 mm longi; flores 4-meri (raro 5- 
meri), 4.5-5 mm longi; calyx 2.5 mm longus et 
3 mm latus, glaber, lobis 3 vel 4, triangularibus, 
acutiusculis, 1—1.5 mm longis, corolla anguste 
campanula ta, 4.5-5 mm longa, externe minute 
strigillosa, intus glabra, ad medium in lobis 4 
(raro 5) divisa, lobis lanceolatis, 2.5 mm longis 
et 1 mm latis, acutis, valvatis, carinatis, eras- 
sis, marginibus apiculoque inflexis; stamina 8 
(raro 10), sub disco inserta, 1.5 mm longa, 
glabra, antheribus 1 mm longis; discus annu- 
laris, 0.7 mm longus, 2 mm diametro, glaber, 
margine crenatus; ovarium ovoideum, 1.5 mm 
longum et 1.3 mm diametro, sparse pilosum, 4- 
loculare (raro 5-loculare) ovulis in loculis 2; 
stylus brevis; stigma capitatum, 4-lobulatum 
(raro 5-lobulatum). Fructus deest. 

Large tree 25 m tall and 50 cm in trunk diam- 
eter, in wet tropical forest. Wood used for 
lumber. "Quemapecho." 

Esmeraldas: San Lorenzo, altitude about 
10 m, April 20, 1943, Little 6292 (Forest Serv- 
ice no. 96876; U. S. Nat. Herb. no. 1877653, 

Six species of Tetragastris Gaertn. are recog- 
nized in the recent monograph by Swart (Rec. 
Trav. Bot. Neerland. 39: 403-419, illus. 1942), 
and another was published in 1939. This generic 
name is retained here over the technically valid, 
older generic name Hedwigia Sw. (1788), which 
is not in use. Hedwigia [Ehrh. 1781] Beauv. 
(1805) should be made a nomen conservandam 
for a monotypic genus of mosses (Little, 
Bryologist 46: 114-115. 1943). The genus is 
distributed from Brazil and British Guiana to 
the West Indies and Central America and now 

Mar. 15, 1948 little: new species of trees from western Ecuador 


Fig. 11. — Symplocos ecuadorensis, £X; flowers, IX; fruit, IX. Fig. 12. — Humiria procera, in- 

cluding fruit and two old fruits showing bony endocarp, all f X. Fig. 13. — Neosprucea pedicellata, 
| X ; flower f X. Fig. 14. — Erythrochiton carinatus, f X ; carpels of fruit, 1 X. 



is reported for the first time from Ecuador and 
the Pacific slope of South America. 

Tetragastris varians is easily distinguished by 
the broadly elliptic leaflets, which vary in 
number. Leaves at the ends of twigs have only 
3 leaflets or 1, but apparently some leaves have 
more leaflets. Other species of the genus have 
narrower leaflets, 5 (rarely 3) to 11 in number. 
In the new species the calyx is glabrous, the 
corolla is minutely strigillose outside and 
glabrous inside, and the ovary is sparsely pilose. 
Specimens of the type collection in the U. S. 
National Herbarium and Forest Service 
Herbarium are terminal portions of twigs with 
both trifoliolate and unifoliolate leaves. How- 
ever, the type bears, at a distance of 30 cm 
from the apex of the twig, the base of a much 
longer pinnate leaf with petiole 21 cm long and 
with a single leaflet attached. 

Guarea polymera Little, sp. nov. 
Fig. 4 

Sect. Euguarea C. DC. Arbor 12 m alta, 
myrmecophila, trunco 15 cm diametro. Ramuli 
minute et dense brunneo-glanduloso-tomentu- 
losi. Folia pinnata; petioli 6-7 cm longi, supra 
canaliculati ; rhaches ca. 26 cm longi; foliolae 
6- vel 7-jugae, breviter (4-5 mm) petiolulatae; 
laminae oblongo-lanceolatae, base acutae, apice 
acuminatae, membranaceae, 20-23 cm longae, 
6-7 cm latae, supra glabrae, subtus incon- 
spicue et minute cinereo-puberulae, nervis 
lateralibus utrinque 12-15, ascendentibus et 
margine curvatis, his cum Costa supra fere 
planibus. subtus prominentibus. Paniculae 
axillares 6-12 cm longae, 3-6 cm latae, ramulis 
minute glanduloso-tomentulosis, bracteae tri- 
angulares, acutae, 2-3 mm longae; pedicelli 
clavellati 5-15 cm longi. Flores albidi, fra- 
grantes, (5) 6- vel 7-meri, calyx late cupu- 
latus, 3 mm longus, 7 mm latus, leviter 10-16- 
lobatus, lobulis 0.5 mm longis, mox 3-4-fissus 
fere ad basin, externe brunneo-tomentulosis, 
intus glaber. Petala 5-7, valvata, lanceolata, 
acuta, leviter involuta, 10 mm longa, 3 mm 
lata, externe dense albido-sericeo-tomentulosa, 
intus glabra; tubus stamineus cylindricus, mar- 
gine crenulatus, 7 mm longus, 4 mm diametro, 
glaber; antherae 10-14, inclusae, sub margine 
sessiles, 1.5 mm longae; gynophorum 1 mm 
altum, glabrum, apice annulum glandulosum 
gerens, 4 mm diametro; ovarium ovoideum, 

sulcatum, dense sericeo-tomentulosum, 3 mm 
longum, 2.5 mm diametro, 6-8-loculare, locu- 
lis 2-ovulatis, ovulis superpositis; stylus cras- 
sus, strigillosus, 3 mm longus, 1 mm diametro; 
stigma crassum discoideum, glabrum, 2 mm 
diametro. Capsula fere matura globosa, rugu- 
losa, brunnea, 11 mm diametro, 6-8-loculare; 
semina in loculo 2 superposita, atra, ca. 3 mm 

Tree 12 m tall and 15 em in diameter, said to 
become larger. Twigs hollow and inhabited by 
ants. Wet tropical forest. "Pialde macho." 

Esmeraldas: Playa de Oro, altitude 65 m, 
April 30, 1943, Little 64.04 (Forest Service no. 
98254; U. S. Nat. Herb. no. 1877588, type; 
with wood sample). 

This specimen is readily associated with the 
group of a few species in the large genus 
Guarea Allem. ex. L. and Sect. Euguarea C. 
DC. having ovaries and capsules with more 
than 5 cells (6 to 12). The flowers are 6- or 7- 
merous, rarely 5-merous, and the ovaries are 
6-8-celled. Petals are 5 to 7, instead of the usual 
number in this genus, 4 or 5. The wood sample 
was identified by S. J. Record as Guarea. Other 
species with approximately the same number of 
carpels are G. purusana C. DC, of Amazonian 
Brazil, with 6-9-celled ovary but much larger 
fruit and dense brownish pubescence, and G. 
grandifolia DC, of Guiana, 7-celled but other- 
wise 4-merous. This Ecuadoran collection con- 
tains also specimens of G. syringoides C. H. 
Wright (?) and G. trichilioides L. 

Perhaps Guarea Allem. ex L. (Mant. PL 2: 
150. 1771) should be made a nomen conservan- 
dum over Elutheria P. Br. (Civ. Nat. Hist. 
Jamaica 369. 1756). The latter was referred to 
Guarea by Fawcett and Rendle (Fl. Jam. 4: 
215. 1920) and cited by Harms (in Engler and 
Prantl, Naturl. Pflanzenfam. ed. 2, 19bl: 129. 
1940). Rendle (Journ. Bot. 50: 129. 1912) con- 
tended that it was unnecessary to conserve 
names against earlier names by Patrick Browne 
observing that Browne published descriptions 
of species only, not genera. Whether Browne's 
descriptions in monotypic genera, such as 
Elutheria, would be valid as combined generic 
and specific descriptions (art. 43) is uncertain 
because binomial nomenclature was not 
adopted. Nevertheless, several generic names 
of Browne having priority have formally been 
made nomina rejicienda. 

Mar. 15, 1948 little: new species of trees from western Ecuador 


Trichilia floribunda Little, sp. nov. 
Fig. 10 

Sect. Moschoxylum (A. Juss.) C. DC. Arbor 
10-20 m alta vel major, trfcnco 20 cm diametro. 
Ramuli minute strigillosi. Folia alterna, magna, 
imparipinnata; petioli 8-11 cm longi; rhaches 
30-35 cm longae; foliolae 11-14, alternae; 
petioluli 5-12 mm longi; petioli, rhaches, et 
petioluli minute strigillosi; laminae foliolarum 
oblanceolatae, (11) 25-33 cm longae et (5) 7-10 
cm latae, infimae minores, basi acutae, apice 
acutae, integrae, subcoriaceae, supra glabres- 
centes, subtus minute et sparse puberulentae 
vel glabrescentes, nervis lateralibus utrinque 
(9) 20-25, ascendentibus, marginem versus 
curvatis, his cum costa supra fere planibusvel 
leviter impressis, subtus prominentibus. Pani- 
culae axillares, corymbiformes, multiflorae, 
20-30 cm longae, fere 20 cm latae, ramulis 
strigillosis; pedicelli tenues, 1-2 mm longi; 
flores flavo-albidi, fragrantes; alabastrum 
oblongo-ovoideum, 6 mm. longum et 2-2.5 mm 
latum; calyx patelliformis, brevissimus, sub- 
integer vel leviter 4- vel 5-lobulatus, 0.5 mm 
longus et 2.5 mm latus, externe strigillosus, 
intus glaber; corolla in alabastro 6 mm longa 
et 2-2.5 mm diametro, externe dense cinereo- 
strigillosa, intus glabra; tubus cylindricus, 4- 
lobatus (raro 3-lobatus) fere ad medium, lobis 
lanceolatis, obtusis, cucullatis, valvatis, leviter 
involutis; tubus stamineus urceolatus, 4 mm. 
longus, apice 6-8 denticulatus (denticulis 1 
mm longis), externe glabrescens vel parce ap- 
presso-pilosus, intus hirtellus; antherae 6-8, 
vulgo 7, inter denticulos insertae, exsertae, 1 
mm longae; ovarium ovoideum, cum gyno- 
phoro 1.5 mm longum et 1 mm latum, dense 
strigillosum, 2- vel 3-loculare, loculis 2-ovu- 
latis; stylus tenuis, inferne strigillosus, 1.5 mm 
longus; stigma clavatum. Fructus deest. 

Tree 10-20 m tall and 20 cm in diameter, re- 
ported to become a large tree of forest canopy. 
Wood said to be used for axe handles. Possibly 
of ornamental value because of the abundant 
small, yellowish-white, fragrant flowers. Wet 
tropical forest of northwestern Esmeraldas. 

Esmeraldas: San Lorenzo, altitude 10 m, 
April 21, 1943, Little 6323 (Forest Service no. 
96852; U. S. Nat. Herb. no. 1877646, type); 
Borb6n, altitude 15m,April26, 1943, Little 6372 
(Forest Service no. 98283; with wood sample). 

The very numerous flowers in this species of 
Trichilia P. Br. have a saucerlike calyx very 
slightly 4- or 5-lobed, and the 4-lobed corolla 
united more than half its length. This species 
has relatively large leaves and leaflets for the 
genus. S. J. Record named the wood sample as 
Trichilia. Also represented in this collection 
are T. hirta L and T. macrcphylla Benth. 

Belotia australis Little, sp. nov. 
Fig. 1 

Arbor 18 m alta vel major; trunco 35 cm. 
diametro. Cortex glaber, albidus. Ramuli et 
petioli minute dense stellato-puberuli, grisei. 
Petioli 8-15 mm longi. Laminae ellipticae vel 
lanceolatae, 9-14 cm longae, 3-6 cm latae, 
basi obtusae vel acutae, apice acuminatae, 
acumine 1-1.5 cm longo, glanduloso-serrulatae, 
3-nervatae, chartaceae, discolores, supra viri- 
des, minute sparse stellato-puberulae, subtus 
griseae, minute dense stellato-puberulae, pilis 
paucis stellatis majoribus praeditae. Inflores- 
centia cymosa axillaris, 3-6 cm longa, ca. 10- 
flora; pedicelli 3-6 mm longi; bracteae lanceo- 
latae, 2-4 mm longae, interdum 2- vel 3-den- 
tatae; pedunculus et pedicelli bracteaeque 
minute dense stellato-puberulae, griseae; ala- 
bastra ca. 8 mm longa; flores 10-11 mm longi; 
sepala 5, rubella, lanceolata, 9-11 mm longa, 
2.5 mm lata, margine involuta, apice obtusa, 
cucullata, 5-nervata, externe minute dense 
stellato-puberula, intus fere glabra, pilis paucis 
stellatis praedita; petala 5, ligulata, 8-9 mm 
longa, 1.5 mm lata, apice 2-4-dentata, 5-ner- 
vata, externe sparse pubescentia, intus basi et 
circum nectarium stellato-pilosa, ceterum gla- 
bra; androgynophorum 1 mm altum, apice 
discum dense stellato-pilosum 3 mm diametro 
ferens; stamina 15, 2-3 mm longa; filamenta 
deorsum stellato-pilosa; antherae suboribcu- 
lares, 0.3 mm longae; ovarium ovoideum, 3 mm 
longum, 2.5 mm diametro, dense pilosum, 2- 
loculare, ovulis paucis; stylus 3-4 mm longus, 
basi stellato-pilosus; stigma 4-lobatum, lobis 
laciniatis, 0.5 mm longis. Capsula bilocularis, 
valde obcompressa, loculicida, brunnea, 15 
mm longa, 20 mm lata, 4 mm crassa, basi ro- 
tundata, apice truncata, sparse puberula, 
pilis minutis stellatis et pilis stellatis multo 
longioribus interspersis, stylo persistente cus- 
pidato 3 mm. longo coronata. Semina desunt. 

Tree 18 m tall or larger, and 35 cm in diame- 



ter. Bark smooth, whitish. The light-weight 
wood is sometimes mixed with that of balsa 
(Ochroma lagopus Sw.) and substituted for the 
latter. Scattered in cutover wet tropical forest 
and fields at San Lorenzo but not common. 
"Chillarde." Also improperly called "balsa." 
Esmeraldas: San Lorenzo, altitude about 

10 m, April 19, 1943, Little 6273 (Forest Service 
no. 96826; U. S. Nat. Herb. no. 1858843, type; 
with wood sample). 

In the genus Belotia A. Rich., T. A. Sprague 
(Kew Bull. 1921: 270-278. 1921) distinguished 

11 species, including 6 new, distributed from 
southern Mexico through Central America to 
Panama and in Cuba and Santa Lucia in the 
West Indies. A. A. Bullock (Kew Bull. 1939: 
517-521. 1939) later examined additional ma- 
terial and reduced the number of species to 9. 
Sprague predicted that this genus might be 
found later in the Pacific coast forests of Co- 
lombia, and this collection represents a further 
southward extension of Belotia to Ecuador. 

Belotia australis is related to B. panamensis 
Pittier (B. macrantha Sprague) of Panama. 
Pubescence in the former is finer, more minute, 
and gray, while the latter has the twigs and 
branches of inflorescence coarsely ferrugineous 
tomentose. The leaves of B. panamensis usu- 
ally are larger, rounded or obtuse at base, and 
with teeth more prominent. The under surface 
of the leaves in B. australis has finer pubescence 
with the larger hairs of the upper tier scattered 
and fewer. The flowers of B. panamensis are 
slightly larger. Capsules in both species are 
similar in size and shape, but those of B. pana- 
mensis are much more densely stellate pubes- 
cent and have shorter styles less than 3 mm 

Sprague noted that the common name 
"balsa," which is applied to B. australis, is used 
also in British Honduras for B. campbellii 

Matisia alata Little, sp nov. 
Fig. 16 

Arbor 20 m. alta, trunco 20-30 cm diametro. 
Ramuli et petioli minute stellato-tomentosi, 
pilis flavis, radiis numerosis, brevissimis, 
glandulosis, demum glabrescentes. Folia magna 
petiolata, petiolis 2-5 cm longis; laminae 
oblongo-obovatae, 15-38 cm longae, 8-18 cm 
latae, basi subrotundatae vel leviter cordatae, 
apice acutae, margine integrae vel obscure un- 

dulatae, penninervis, nervis lateralibus 9-12 
utrinque latere, prope marginem arcuatis et 
anastomosantibus, subtus elevatis, venulis 
prominente reticulata, supra glabratae, subtus 
sparse et minute stellato-puberulae. Flores 
solitarii, oppositifolii, 2 vel 3 apice ramulorum, 
magni, 7 cm longi; pedicellus crassus, curvatus, 
fulvus, 4-7 cm longus, 3 mm diametro; calyx 
anguste campanulatus, 45 mm longus, 22 mm 
diametro cum alis, basi attenuatus, crassus, 
plus minusve succulentus, ochraceus, dense et 
minute stellato-.tomentosus, pilis glandulosis, 
intus dense sericeus, lobis 5 brevibus, acutis, 
3-5 mm longis, alis 10, longis, angustis, usque 
ad 6 mm latis; petala 5, ochracea, 70 mm longa, 
basi connata tubo 2 mm, limbo spatulato vel 
obovato, obtuso, ca. 32 mm longo et 15-18 mm 
lato, externe dense stellato-tomentoso, intus 
glabro; columna staminea longe exserta, leviter 
curvata, 5 cm longa, 5-lobata, lobulis leviter 
pubescentibus, 10-16 mm longis, antherarum 
loculis 4-6 linearibus; ovarium 5-lobatum, 5- 
loculare, 3 mm longum, stellato-tomentosum; 
stylus 60 mm longus, 10 mm exsertus, curva- 
tus, stellato-tomentosus; stigma capitatum, 2 
mm longum. Fructus globosus, succulentus, 
ochraceus, magnus, 8-10 cm diametro, calyce 
accrescente 6-7 cm longo, cum alis 4-10 mm 
latis, suffultus. 

Tree 20 m tall and 20-30 cm in trunk diam- 
eter. It is said that the large, fleshy fruits are 
edible. Observed also at San Lorenzo, Esmeral- 
das. Wet tropical forest of Esmeraldas and 
Pichincha, northwestern Ecuador. "Sapote," 

Esmeraldas: Quininde, altitude 65 m, April 
12, 1943, Little 62U (Forest Service no. 96751; 
U. S. Nat. Herb. no. 1858828, type). Pinch- 
incha: between Santo Domingo de los Colora- 
dos and Quininde, April 7, 1943, Little 6197 
(Forest Service no. 96867); April 8, 1943, 
Little 6200 (Forest Service no. 96785; in fruit). 

More than 20 species of the genus Matisia 
H. B. K. have been described, a third of these 
in 1945-1946. This genus of small to medium- 
sized trees ranges from Brazil to Colombia and 
Ecuador and north to Panama and Costa Rica. 
Some of the species apparently are of local 
range. In this Ecuadorian collection are M. 
cordata Humb. & Bonpl., a cultivated species, 
M. coloradorum R. Benoist, an Ecuadorian 
species found again at the type locality, and 
two species proposed here as new. 

Mar. 15, 1948 little: new species of trees from western Ecuador 


Fig. 15. — Clusia poly stigma, including bud and immature fruit, all |X. Fig. 16. — Matisia alata, 

| X; flower, |X. Fig. 17. — Clusia plurivalvis, including separate staminate flowers and fruits, 

all ^ X. Fig. 18. — Matisia grandifolia, £ X; flower, about 1 X . 



Matisia alata, one of the few species with 
pinnately veined leaves, is distinguished from 
other species of the genus by the prominent 
wings on the calyx, to which the specific name 
refers, and by the large flowers, which are 
orange or brownish in color. The 10 conspicu- 
ous wings of the calyx tube are longitudinal, 
narrow, slightly curled projecting ridges, start- 
ing at the base of the calyx and reaching a 
maximum width of 6 mm. near their rounded 
ends of the apex of the calyx. The 5 wings oppo- 
site the calyx lobes are larger, while the alter- 
nate wings are smaller and shorter or may be 
almost absent. 

Matisia grandifolia Little, sp. nov. 
Fig. 18 

Arbor parva, 5 m alta, trunco 5 cm diametro. 
Ramuli et petioli dense et minute stellato-to- 
mentosi, pilis multiradiatis, glandulosis. Folia 
magna, petiolata, petiolis 1-1.5 cm longis, 4 
mm diametro; laminae late ellipticae, 50-60 cm 
longae, ca. 28 cm latae, basi rotundatae, apice 
acuminatae, margine integrae vel obscure un- 
dulatae, penninervis, nervis lateralibus 12-14 
utrinque latere, prope marginem arcuatis et 
anastamosantibus, subtus elevatis, venulis 
prominente reticulars, supra glabrae, subtus 
minute stellato-puberulae. Flores pauci, soli- 
tarii, oppositifolii, 3-3.5 cm longi; pedicellus 
brevis, leviter curvatus, 5-7 mm longus; brac- 
teis 3, basi floris, subulatis, 3-6 mm longis; 
calyx anguste campanulatus, fulvus, 22-25 mm 
longus, 8 mm diametro, uno latere 10-12 mm 
fissus, lobis plerumque cohaerentibus, saepe 
2-4, sinubus 2-7 mm longis, alis 10 longis, 
angustis, 1 mm latis, dense et minute stellato- 
tomentosus, pilis glandulosis, intus dense seri- 
ceus, petala 5, in sicco rubida, anguste oblan- 
ceolata, apice obtusa, 30-34 mm longa, 4-5 mm 
lata, externe apice appresso-stellato-tomentosa; 
columna staminea 17-20 mm longa, apice le- 
viter pubescens, 4- vel 5-lobata, lobulis 3-9 mm 
longis, antherarum loculis ellipticis usque ad 8; 
ovarium conicum, 5-loculare, 3-4 mm longum, 
stellato-tomentosum; stylus 24 mm longus, 3 
mm exsertus, stellato-tomentosus; stigma 5- 
lobatum, lobulis obtusis, fere 2 mm longis. 
Fructus deest. 

Small tree 5 m tall and 5 cm in diameter, 
said to become larger. One tree seen in cacao 
plantation, wet tropical forest area. "Peni- 

Pichincha: Santo Domingo de los Colora- 
dos, altitude 560 m, April 7, 1943, Little 6146 
(Forest Service no. 96782; U. S. Nat. Herb. no. 
1858834, type). 

This species is characterized by unusually 
large, pinnately veined leaves, by the short 
pedicels, and by the calyx split on one side, 
with 10 narrow wings or ridges. 

Clusia plurivalvis Little, sp. nov. 
Fig. 17 

Arbor dioica, recta, extensa, 15 m alta, 
trunco 30 cm diametro. Cortex fuscus, leviter 
asper, fissus, latice alba. Ramuli crassi, teretes. 
Folia petiolata; petioli 1.5-4 cm longi, 4-7 mm 
lati, leviter alati; laminae ovales vel obovatae, 
(8) 12-24 cm longae, (5) 8-14 cm latae, basi 
obtusae vel acutae, apice rotundatae, valde 
coriaceae, nervis lateralibus numerosis, paral- 
lels, sub angulo ca. 60-70° adscendentibus. 
Inflorescentia terminalis, capitata, pauciflora 
pedunculo crasso, 2.5-4 cm longo, 6 mm dia- 
metro, floribus sessilibus. Flores masculi magni, 
5 cm diametro; bracteae calycinae rotundae, 
4-9 mm longae, 8-13 mm latae, obtusae, 
carinatae; sepala 4-6, biseriata, late rotundata, 
obtusa, concava, coriacea, 2 inferiora minora 
6-10 mm longa et 15 mm lata, superiora 10-15 
mm longa et 12-20 mm lata; petala 5, obovata, 
28-32 mm longa, 20-25 mm lata, obtusa, in 
vivo alba, basi incarnata vel ochracea; stamina 
in disco convexo 5-6 mm alto et 10-13 mm dia- 
metro inserta, numerosissima, spissa, fere 
sessilia, antheris 1.5 mm longis, longitudina- 
liter dehiscentibus; staminodia in globo re- 
sinoso 5 mm diametro apice disci aggregata. 
Flores feminei desunt; sepala in fructu per- 
sistentia, similia sepalis florium masculorum; 
staminodia nulla. Capsula succulenta depresso- 
globosa, 17 mm alta et 21 mm diametro, 13- 
16-locularis et 13-16-valvatis; stigmata 13-16, 
peltata, radiata, sessilia, triangularia, plana. 
8 mm longa et 2-3 mm lata, atra, in disco 17 
mm diametro; semina in loculis 2. 

Large, erect, widely spreading tree 15 m tall 
and 30 cm in trunk diameter. Bark dark brown, 
slightly rough, with white latex. Common and 
with another, probably undescribed species 
of Clusia (Little 6769) a dominant tree species 
of the dry mountain forest in ravines at Huigra. 
Common name "lamai," according to Rose. 

Chimborazo: Huigra, altitude 1300 m, July 
17, 1943, Little 67 68 A (Forest Service no. 

Mar. 15, 1948 little: new species of trees from western Ecuador 


98525; IT. S. Nat. Herb. no. 1857089, type; 
with fruits). Little 6768 (Forest Service no. 
95852; with staminate flowers). Little 6768 B 
(Forest Service no. 98526; with staminate 
flowers). Huigra, Aug. 19, 1918, J. N. Rose and 
George Rose 2223 (U. S. Nat. Herb. no. 1021907; 
with fruits and staminate flowers but also 
containing larger, detached, sectioned fruits 
of a different species). 

Rose and Rose 2223 bears the note, " 'Not 
at Kew' N. L. B[ritton]. Sept. 1920." Clusia 
plurivalvis is not readily placed in the proper 
section in the latest summary of Clusia L. 
by Engler (Nattirl. Pflanzenfam. ed. 2, 42: 

199-204. 1925). It seems to be in subgenus 
Thysanoclusia Vesque and related to section 
Euclusia Planch. & Triana, which differs in 
having the connectives elongate and pointed. 
In this species the numerous stamens are 
crowded and almost sessile in a convex disc, 
which bears a central resinous mass of stami- 
nodia. The number of carpels in the ovary in 
this species, 13-16, is very high. Of the species 
in which the number of carpels is known, only 
a few, such as C. flava Jacq. with 12-14 car- 
pels, C. cuneata Benth. with 16, and the new 
species described below have more than 12 

Fig. 19. — Aspidosperma datum, ?X; fruit, |X. 



Clusia polystigma Little, sp. nov. 
Fig. 15 

Epiphyta magna, scandens, arborea, dioica, 
10 m alta, trunco 10 cm diametro. Ramuli 
crassi, teretes. Folia subsessilia, obovata, 15-20 
cm longa, 8-11 cm lata, basi angustata et sub- 
petioliformes, apice obtusae, valde coriaceae, 
nervis lateralibus numerosis, parallelis, sub 
angulo ca. 50° adscendentibus. Flores masculi 
desunt. Inflorescentia feminea terminalis, race- 
mosa, pauciflora, rhache 6 cm longa, bracteata, 
bracteis oppositis, late ovatis, 5-8 mm longis 
et 6-10 mm latis; pedicelli 3-6 mm longi; 
bracteae calycinae 2 late rotundatae, 3 mm 
longae et 5-6 mm latae, obtusae, crassae; 
sepala 4-6, biseriata, late rotundata, obtusa, 
coriacea, minora 2 inferiora, 7-11 mm longa et 
9-12 mm lata, 2-4 superiora 15-16 mm longa 
et 12-18 mm lata; petala 6 vel 7, obovata, 
32-34 mm longa et 22-24 mm lata, obtusa, in 
vivo albida, leviter incarnato-tincta; cupula 
staminodialis 3-4 mm longa, 12 mm diametro, 
apice leviter lobata; ovarium cylindricum, 4 
mm longum, 6 mm diametro, sulcis longitudi- 
nalibus et loculis 16-19, loculis multiovulatis; 
stigmata 16-19, radiata, sessilia, anguste tri- 
angularia, 5-6 mm longa et 1.5 mm lata, atra, 
connata in annulo 4-5 mm longo et 8-9 mm 
diametro. Capsula succulenta immatura sub- 
globosa, 18 mm longa et 16 mm diametro, sul- 
cata, stigmatibus sessilibus coronata. 

Large woody vine 10 m tall and 10 cm in 
trunk diameter, epiphytic on a tree of Ficus sp. 
nov. (Little 6356) left in a clearing, wet tropical 
forest. "Matapalo." 

Esmeraldas: San Lorenzo, altitude about 
10 m, April 23, 1943, Little 6355 (Forest Service 
no. 98245; U. S. Nat. Herb. no. 1857080, type). 

In the absence of staminate flowers, this 
species of Clusia L. cannot be definitely placed 
as to section. When more material is available, 
this species perhaps may be included in sub- 
genus Thysanoclusia Vesque, section Euclusia 
Planch. & Triana, according to the summary 
by A. Engler (Natiirl. Pflanzenfam. ed. 2, 21: 
199-204. 1925). The large, 6- or 7-petaled 
flowers suggest affinities with this section. This 
species is readily distinguished by the very high 
number of carpels, 16-19, apparently more 
than have been recorded previously in this 

Neosprucea pedicellata Little, sp. nov. 
Fig. 13 

Arbor parva, 10 in alta, trunco 20 cm dia- 
metro, ramulis teretibus griseis, minute ap- 
presso-pubescentibus. Folia petiolata, pe- 
tiolis appresso pubescentibus, 12-16 mm longis; 
laminae ellipticae, 10-20 cm longae, 6-10 
cm latae, membranaceae, basi subrotundatae, 
apice attenuatae, margine remote obtuseque 
glanduloso-serratae, e basi distincte tripli- 
nerviae, glabrae nervis subtus minute paullo 
appresso-pubescentibus exceptis. Racemus la- 
teralis laxus divergens, pauciflorus (2 vel 3), 
5-6 cm longus; rachis et pedicelli subangulati, 
minute denseque tomentosi; pedicelli 25-30 
mm longi; flores magni, albidi, sepala et petala 
similia, valvata, persistentia, utrinque breviter 
denseque appresso-tomentosa; sepala 4, basi 
leviter connata, ovata, 16-17 mm longa, 
5-7 mm lata, acuminata, 5-8-nervata; petala 
4, ovata, acuta, carinata, 15-17 mm longa; 
stamina numerosa; filamenta filiformia, 2 m m 
longa; antherae linearae, hirsutae, 5 mm 
longae; receptaculum dense hirtum; ovarium 
globosum, fere glabrum, 5 mm diametro, 
semi-5-7-loculare, placentis parietalibus pro- 
minentibus; stylus crasse filiformis, 9 mm lon- 
gus; stigmate truncate Fructus deest. 

Small tree about 10 m. tall and 20 cm. in 
diameter, wet tropical forest. 

Pichincha: Santo Domingo de los Colorados 
altitude about 560 m, April 2, 1943, Little 6152 
(Forest Service no. 96909; U. S. Nat. Herb. no. 
1858861, type). 

Neosprucea Sleumer (Notizbl. Berl. 14: 47. 
1938), a monotypic genus of the Flacourtiaceae, 
was based upon A", gr audi flora (Spruce) Sleu- 
mer (Notizbl. Berl. 44: 47. 1938), originally 
described as Banara grandiflora Spruce (Journ. 
Linn. Soc. Bot. 5. Suppl. 2: 93. 1861). The type 
was collected at Tarapoto, Peru, by Spruce, 
and the species now is known from the Atlantic 
slope in Brazil, Colombia, and Peru. The sec- 
ond species, N. pedicellata, represents an ex- 
tension of range of the genus to Ecuador and 
possibly the first record on the Pacific slope. 

A r . pedicellata is a distinct species readily 
separated from the type species by the follow- 
ing characters: leaves membranaceous (instead 
of chartaceous) ; leaves mostly larger, broader, 
and rounded at base; petioles slightly longer; 

Mar. 15, 1948 little: new species of trees from western Ecuador 


greatly elongated pedicels 25-30 mm long; 
and larger flowers. Though N. grandiflora was 
described by Sleumer with a spikelike raceme 
and pedicels only 2 mm. long, Mutis' specimens 
from Colombia have pedicels 10 to 20 mm. 

Symplocos ecuadorensis Little, sp. nov. 
Fig. 11 

Subgenus Eusymplocos Brand, sect. Symplo- 
castrum Brand, subsect. Ciponimastrum Brand. 
Arbor 15 m alta, trunco 20 cm diametro. Ra- 
muli leviter hirsuti. Folia petiolata, petiolis 
hirsutis, 5-8 mm longis; laminae oblanceolatae, 
5-10 cm longae, 2.5-4 cm latae, basi cuneatae, 
apice abrupte acuminatae, remote crenulatae, 
subcoriaceae, lucidae, supra glabrae, subtus 
parce hirsutulae, costa et venis primariis subtus 
prominentibus. Paniculae axillares, 1-2 cm. 
longae, rachibus hirsutis, bracteis ovatis, hir- 
sutis, ciliatis, 1 mm longis; flores 7-8 mm longi, 
subsessiles; calyx 2-2.5 mm longus, hypanthio 
campanulato ca. 1 mm longo, glabro, lobis 5, 
late ovatis, ciliatis, 1-1.5 mm. longis; corolla 
alba, 6-7 mm longa, tubo ca. 3 mm longo, lobis 
5, ellipticis, obtusis, apice ciliolatis, 3-4 mm 
longis et 2 mm latis; tubus stamineus tubo 
corollae adnatus, 1 mm longus; stamina 28-32 
triserialia; filamenta complanata, 0.2-2.5 mm 
longa; ovarium fere omnino inferum, 1.5 cm. 
altum, apice pilosum, 3- vel 4-lobulatum. Fruc- 
tus baccatus, cylindricus, lobis calycinis et 
basi styli coronatus, 10 mm. longus, 5 mm 
diametro, 3- vel 4-locularis. 

Tree 15 m tall and 20 cm in diameter. Two 
trees were seen in a pasture, cleared wet trop- 
ical forest. 

Esmeraldas: San Lorenzo, altitude about 
10 m, April 22, 1943, Little 6349 (Forest Serv- 
ice no. 98370; U. S. Nat. Herb. no. 1858685, 

Though the genus Symplocos Jacq. is widely 
distributed with about 300 species in tropical 
and subtropical America, Asia, and Australia, 
most species have restricted ranges. This ma- 
terial from the poorly explored Ecuadorian 
province of Esmeraldas could not be assigned 

to any of the species from northwestern South 

Aspidosperma elatum Little, sp. nov. 
Fig. 19 

Arbor magna, 35 m alta; trunco 1 m dia- 
metro, profunde sulcato et angulato. Cortex 
fere laevis, leviter fissus, albido-griseus. Ra- 
muli subteretes, puberuli. Folia alterna, non 
congesta, internodiis 4-7 cm longis, petiolata, 
petiolis fusco-puberulis, 1 cm longis; laminae 
magnae, oblongo-oblanceolatae, chartaceae, 
22-26 cm longae, 7-9 cm latae, basi subrotun- 
datae, apice acutae vel obtusae, fere glabrae, 
supra virides, subtus glaucescentes, costa 
leviter puberula, supra canaliculata, subtus 
elevata, nervis lateralibus utrinque 13-16, fere 
rectis, prope marginem leviter arcuatis, subtus 
elevatis, secundariis inconspicuis. Inflorescen- 
tiae desunt. Mericarpia magna, compressa, 
suborbicularia, 10-12 cm longa, 9-10 cm lata, 
2 cm crassa, externe dense fusco-tomentosa, 
pericarpio lignoso, 5-8 mm crasso; semina 
plano-compressa, suborbicularia, 9 cm longa, 
7 cm lata, alis latis, papyraceis. 

Large tree of forest canopy, to 35 m tall and 
1 m in trunk diameter. Easily recognized in the 
field by the odd, fluted trunk with deep, 
branching, vertical grooves. Bark smoothish, 
slightly cracked, whitish-gray. Wet tropical 
forest at Pichilingue. "Naranjo de monte." 

Los Rios: Pichilingue, altitude about 45 m, 
May 25, 1943. Little 6517 (Forest Service no. 
98223; U. S. Nat. Herb., type). 

The genus Aspidosperma Mart. & Zucc. is 
widely distributed in tropical America from 
Mexico and the West Indies to Brazil and Ar- 
gentina, centering in Brazil and represented in 
northwestern South America by only a few 
species. Four species were recorded from Col- 
ombia by Standley (Trop. Woods 36: 13-20. 
1933). The herbarium material of Aspido- 
sperma elatum, which consists of foliage and 
detached fruits, seems sufficiently distinctive, 
as the leaves and fruits are larger than in most 
species. Apparently this is the second species of 
Aspidosperma recorded from Ecuador. 



HERPETOLOGY. — A collection of salamanders from Mount Rogers, Virginia. 1 
Richard L. Hoffman and Hubert I. Kleinpeter. (Communicated by 
Herbert Friedmann.) 

Although White Top Mountain, Va., is 
well known to students of salamander dis- 
tribution, and is subjected to periodic dep- 
redation by different collectors; its ad- 
jacent neighbor to the northeast, Mount 
Rogers, has apparently escaped the atten- 
tion of most herpetologists. The absence of 
any particular road to the summit of MoUnt 
Rogers may be partly responsible for this 

When, in the summer of 1947, we had 
the opportunity to be in southwestern Vir- 
ginia, we avoided White Top believing that 
we could make no particular contribution 
by further depleting its already well-known 
herpetological fauna. Instead a period of 
three days was spent on Mount Rogers. 
Our station was located on the eastern slope, 
in Grayson County, at an elevation of about 
4,500 feet, from which one all-day journey 
was made to the top, as well as many short 
trips in the general region about the camp. 

The lower slopes of the mountain are 
thickly forested with a mesophytic flora 
which might be characterized as a maple- 
hemlock-dogwood association, with com- 
paratively little undergrowth of herbaceous 
plants. Although fir and spruce trees are 
widely scattered, the extensive stands are 
restricted to the uppermost 500 or 600 feet 
of the knob. Here red spruce, Picea rubra 
(DuRoi) Dietr., and southern fir, Abies 
fraseri (Pursh) Poir., seem to comprise the 
entire evergreen forest, the latter being by 
far the more abundant. The ground is com- 
pletely covered by mosses and thick car- 
pets of the wood sorrel, Oxalis acetosella L. 
A striking feature of the evergreen stands 
is the large number of logs and stumps, all 
in approximately the same stage of de- 
composition. The more level parts of the 
top are covered by heavy growths of vari- 
ous ferns. 

From a physiographic standpoint, the 
Balsam Mountains (of which White Top 
and Mount Rogers form the bulk) are char- 

1 Received October 20, 3 947. 

acteristic of the Southern Section of the 
Blue Ridge Province. White Top has previ- 
ously been designated part of the Iron 
Mountains, but the recent Mount Rogers 
Quadrangle (q.v.) of the U. S. Geological 
Survey ranks the Iron and Balsam 
Mountains as separate units. 

Although a rather large number of sala- 
manders was encountered, they represented 
but a few species. The contrast between the 
salamander fauna of White Top and Mount 
Rogers is interesting in that the following 
species reported from the former were not 
found by us on Mount Rogers: Triturus v. 
viridescens, Desmognathus monticola, Ple- 
ihodon c. cinereus, P. g. glutinosus, P. 
yonahlossee, and Pseudotriton ruber nitidus. 
It is difficult to account for the apparent 
absence of several of these species, whereas 
at least one other form which is scarce on 
White Top (Plethodon welleri), is extremely 
common on Rogers. In the case of P. yonah- 
lossee, it is not inconceivable that we col- 
lected above its maximum elevation prefer- 
ences. The lower slopes seemed very favor- 
able for P. glutinosus, but the only Pletho- 
don found there was metcalfi, which was 
present in considerable numbers. 

In addition to the salamanders which 
(with the exception of a small series of a 
strange Desmognathus reserved by the jun- 
ior author for further examination) have 
been donated to the United States National 
Museum, the following amphibians and rep- 
tiles were found on Mount Rogers : Bufo ter- 
restris americanus, Lampropeltis t. triangu- 
lum, and Thamnophis s. sir talis. One speci- 
men of the last named was seen sunning 
itself on a rock at approximately 5,500 feet; 
the others were all seen near our station at 
4,500 feet. 

We take pleasure in indicating our grati- 
tude to Drs. Horton H. Hobbs, Jr., and 
Arnold B. Grobman for help in preparation 
of this paper, and to Dr. Doris M. Cochran 
for courtesies attendant upon our visits to 
the National Museum. 

Mar. 15, 1948 hoffman & kleinpeter: salamanders from mount rogers, va. 107 

Desmognathus fuscus fuscus (Rafinesque) 

Six specimens, which were not presented to 
the Museum, were collected in a large seepage 
area located in a cleared saddle on the northeast 
side of Mount Rogers. This species was not 
represented farther down in the stream em- 
anating from the spring. 

Desmognathus ochrophaeus carolinensis Dunn 

Sixteen individuals, U.S.N.M. nos. 124455- 
70, were collected at elevations from 4,500 to 
5,700 feet. The species seemed to be equally 
common in evergreen and deciduous forests. 

Desmognathus quadramaculatus quadra- 
maculatus (Holbrook) 

Nine specimens, mostly large adults, 
U.S.N.M. nos. 124471-79. On July 2 we col- 
lected in a small but cold and swift stream on 
the eastern slope of the mountain at an eleva- 
tion of about 4,300 feet. This species was es- 
pecially common; most of the specimens found 
being under submerged rocks in rapids. The 
largest specimen included in our series meas- 
ures 78 mm from snout to vent, a size not ap- 
proached by specimens from farther north in 

Desmognathus wrighti King 

Two adults, U.S.N.M. nos. 124480 and 
124634, were collected at an elevation of ap- 
proximately 5,600 feet. They were found in 
rotting logs in company with Plethodon wel- 
leri. Mount Rogers is the northernmost locality 
at which this diminutive salamander has been 

Plethodon metcalfi Brimley 

Eight specimens, U.S.N.M. nos. 124412-18, 
124631. This species shares with D. o. carolinen- 
sis the distinction of being one of the two most 
abundant salamanders on Mount Rogers, and 
the size of our series is in no way indicative of 
the abundance of the animal. In the evergreen 
forests, metcalfi occurred all the way to the very 
summit of the mountain, although less common 
on the top; in and under logs but not in stumps. 

Plethodon welleri Walker 

Eleven specimens, U.S.N.M. nos. 124442- 
49, 124632-33. The size of the series serves to 

indicate the relative abundance of this hereto- 
fore supposedly rare species. Far more were 
collected than were retained. It was not until 
we were several hundred yards within the 
evergreen forest that we found the first welleri, 
but thereafter almost every log or stump ex- 
amined contained one and occasionally two 
specimens. Most frequently, welleri was found 
in decaying, moss-covered stumps, usually in 
the upper portion. Individuals found in logs 
were usually lying with the body in a "U" — the 
end of the tail near the head. Many of them 
remained still and were easily caught. 

One log, nearly at the top, was found to con- 
tain two females with their eggs. This log, seem- 
ingly identical with many others investigated, 
was supported about 18 inches above the 
ground. The females were discovered in small 
crevices in the damp wood, about 6 inches 
apart, an inch under the surface of a layer of 
moss. Both were curled around the eggs, and 
showed some reluctance to leave them. In fact, 
one, on being dislodged, moved back after a 
short time. The larger clutch contained nine 
eggs, which were slightly pear-shaped and 
averaged 4.2 (3.6-5.0) mm in the largest diame- 
ter. We could detect no particular pedicles, and 
the eggs clung together by the adhesive nature 
of the outer envelopes. The smaller clutch in- 
cluded seven eggs, which are arranged in a ring 
of five with one above and below at the center. 
In color, the eggs were a light creamy yellow, 
and no trace of embryos could be discerned 
within. The mothers differed in size in accord- 
ance with the number of eggs, the larger meas- 
uring 47 mm snout to vent (87.5 mm total 
length) and the smaller 43.5 mm snout to vent 
(79.1 mm total length). It is interesting that of 
the many welleri seen, the only two with egg 
masses were in the same log. We wonder if it 
was due to coincidence or to a tendency of the 
females to congregate at a sort of Wochenstube 
where conditions for hatching and growth (not 
apparent to humans) are optimum. 

Mount Rogers is the fourth locality reported 
for this form, and it is probably more abundant 
there than at any other place. Only a dozen or 
so specimens have been taken altogether on 
White Top, thus its abundance on Mount 
Rogers is the more interesting, particularly in 
view of the fact that we covered such a small 



Gyrinophilus sp. 

One larval specimen, U.S.N. M. no. 124487. 
This individual was found in the stream noted 
above under D. q. quadramaculatus and was but 
one of several seen in this and the rivulet 
where D. f. fuscus was secured. We expended 
several hours in a particular attempt to ob- 
tain adults, which seem to have escaped the 
efforts of White Top collectors as well. One 
would expect the adults, from geographic prob- 
ability, to be Gyrinophilus danielsi. An attempt 
which was made to raise the larva to trans- 

formation was unsuccessful. Collectors who 
visit the Balsams in the future should make 
special efforts to obtain adult specimens. 

Eurycea bislineata wilderae Dunn 

Two adults, U.S.N.M. nos. 124484, 124629, 
were taken, of which one was found under a 
rock in a marshy area at about 4,900 feet on the 
east side of the top. The second example was 
discovered in a thick leaf pile along the side of 
the creek mentioned above under D. q. quadra- 

ORNITHOLOGY. — The races of the silver-breasted broadbill, Serilophus lunatus 
(Gould). 1 H. G. Deignan, U. S. National Museum. 

Through the kindness of the authorities 
of the Chicago Natural History Museum 
(C.N.H.M.), the Academy of Natural 
Sciences of Philadelphia (A.N.S.P.), and 
the American Museum of Natural History 
(A.M.N.H.), I have been enabled to add to 
the series of the silver-breasted broadbill in 
the United States National Museum 
(U.S.N.M.) to make a total of 93 specimens. 

From the taxonomic point of view, this 
has proved to be a highly unsatisfactory 
species. While the races may be immedi- 
ately broken into three main groups, in 
accordance with the color of the lores 
(black, rusty, or ashy), further subdivision 
hinges upon such subtle factors as the 
extent and intensity of tones of gray and 
brown and is complicated by a certain 
degree of individual variation in almost any 
given character. Thus, while subspeciation 
appears in all the zoogeographic areas where 
such might be expected, yet it is a matter of 
the greatest difficulty to set forth intelli- 
gible diagnoses, even when these apply, not 
to individuals, but to series. In the diag- 
noses given below, only characters that 
possess subspecific importance will be 
noted; it may be said at once that no con- 
sequential variation has been found in 
measurements or in the complicated pat- 
terns of wing and tail. 

With the understanding that Serilophus 
rubropygius (Hodgson) represents a closely 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received July 18, 

allied but independent species, I find the 
following populations of S. lunatus ap- 
parently worthy of nomenclatorial recog- 
nition : 

1. Serilophus lunatus polionotus Rothschild 

Serilophus lunatus polionotus Rothschild, Bull. 
Brit. Orn. Club 14: 7. Oct. 30, 1903 (Mount 
Wuchi, Hainan). 

Diagnosis. — The lores black; the sides of the 
head and the ear coverts pale ashy ferruginous; 
the forehead pale ashj^ gray, this color changing 
insensibly to the pale ashy ferruginous of the 
crown and nape; the scapulars and upper back 
ashy gray; the lower back rufous, this color 
changing to rufous-chestnut on the rump and 
upper tail coverts; the innermost secondaries 

Range. — Hainan. 

Specimens examined. — Hainan: Mount 
Wuchi (10 males, 3 females). 

2. Serilophus lunatus elisabethae La Touche 

Serilophus lunatus elisabethce La Touche, Bull. 
Brit. Orn. Club 42: 14. Oct. 29, 1921 (Hokow, 
elev. 300 feet, southeastern Yunnan Province, 

Diagnosis. — The lores blackish; the sides of 
the head and the ear coverts pale ferruginous; 
the forehead pale ashy, this color changing 
insensibly to the ferruginous of the crown 
and nape; the scapulars and upper back deep 
ashy brown; the lower back chestnut-rufous, 
this color changing to rufous-chestnut on the 
rump and upper tail coverts; the innermost 
secondaries chestnut-rufous. 

Mar. 15, 1948 deignan: the races of the silver-breasted broadbill 


Range. — The valley of the Red River from 
the Chinese frontier to its mouths and thence 
southward into northern Annam. 

Specimens examined. — Yunnan: Southeast: 
Hokow (1 female); Annam: Thanhoa Province: 
Lunglunh (2 males); Vinh Province: Phuqui 
(1 female). 

3. Serilophus lunatus impavidus, n. subsp. 

Type.— C.N.H.M. no. 90275, adult male, 
collected at Thateng flat. 15°3l' N., long. 
106°22 / E.), Sara vane Province, Bas-Laos, on 
December 6, 1931, by Jean Delacour (original 
number 743). 

Diagnosis. — The lores black or blackish; the 
sides of the head and the ear coverts pale 
ferruginous; the forehead pale ashy, this color 
changing insensibly to the light ferruginous of 
the crown and nape; the scapulars and upper 
back brownish ashy; the lower back chestnut- 
rufous, this color changing to rufous-chestnut 
on the rump and upper tail coverts; the inner- 
most secondaries chestnut-rufous. 

Range. — Bas-Laos (Boloven region). 

Specimens examined. — Bas-Laos : Sara vane 
Province: "Boloven Plateau" (1 male), Tha- 
teng (4 males, 4 females), Ban Kok (1 male), 
Pakse (1 male, 1 female). 

Remarks. — This race is very near to S. I. 
elisabethae, with which it has been heretofore 
confused, but is easily distinguishable, es- 
pecially in series, by the predominantly ashy 
hue of the scapulars and upper back. Certain 
examples of true elisabethae seem to approach 
it in this respect but prove to be less well- 
made skins, in which the grayish bases of the 
feathers are exposed to view. In series also the 
reds of crown, nape, lower back, rump, upper 
tail coverts, and innermost secondaries average 
slightly paler in impavidus than in elisabethae. 

4. Serilophus lunatus aphobus, n. subsp. 

Type.— U.S.N.M. no. 330375, adult male, 
collected on Khao Laem ( a mountain of the 
San Kamphaeng Range at lat. 14°25' N., 
long. 101°30 r E.), at the southwestern corner 
of the eastern Siamese Plateau, on December 
28, 1930, by Hugh M. Smith (original number 

Diagnosis. — The lores blackish rusty; the 
sides of the head and the ear coverts ferrugi- 
nous; the forehead pale ashy, this color changing 
insensibly to the light ferruginous of the crown 

and nape; the scapulars and upper back ashy 
brown; the lower back chestnut-rufous, this 
color changing to rufous-chestnut on the rump 
and upper tail coverts; the innermost second- 
aries chestnut-rufous. 

Range. — Eastern Siam. 

Specimens examined. — Siam: East: Ban Lam 
Thong Lang (1 male), Khao Laem (1 male, 1 

Remarks. — The present form is similar to 
both S. 1. elisabethae and S. I. impavidus, but 
separable from either by having the lores 
blackish rusty instead of black, and the sides 
of the head and ear coverts a rather more 
vivid ferruginous. The color of the scapulars 
and upper back is intermediate between those 
of elisabethae and impavidus. 

A male from Khao Soi Dao and another from 
Ban Bang Phra, localities in extreme south- 
eastern Siam, are near aphobus but have the 
reds slightly deeper in tone. For the present 
they must be left unnamed. 

5. Serilophus lunatus atrestus, n. subsp. . 

Type.— A.M.N.H. no. 143346, adult male, 
collected at Mengting (lat. 23°33' N., long. 
99°05 / E.), western Yunnan Province, China, 
on February 19, 1917, by Roy C. Andrews and 
Edmund Heller (original number 492). 

Diagnosis. — The lores blackish rusty; the 
sides of the head and the ear coverts bright 
ferruginous; the forehead pale ashy, this color 
changing insensibly to the bright ferruginous 
of the crown and nape; the scapulars and 
upper back ashy brown; the lower back chest- 
nut-rufous, this color changing to rufous- 
chestnut on the rump and upper tail coverts; 
the innermost secondaries chestnut-rufous. 

Range. — Northeastern Burma; Northern 
and Southern Shan States; western Yunnan; 
Haut-Laos; Tongking (west of the Black 
River- Red River divide); northern Siam 
(eastern provinces.) 

Specimens examined. — Yunnan: West: 
Mengting (1 male); Southern Shan States: 
Kengtung State: Mong Len (1 male); Siam: 
Northeast: Muang Lorn Sak (1 male); Haut- 
laos: 5° Territoire Militaire: Muong Yo (2 
males, 2 females); Tongking: Laichau Prov- 
ince: Muong Mo (1 male). 

Remarks. — This race is distinguishable in 
series from S. I. aphobus by the more vivid 



tones of the reds of the sides of the head, ear 
coverts, and upper parts. 

From S. I. elisabethae it may be separated by 
having the lores blackish rusty instead of black, 
and by having the reds of the sides of the head, 
ear coverts, and upper parts distinctly lighter 
and brighter. 

6. Serilophus lunatus lunatus (Gould) 

Eurylaimus lunatus Gould, Proc. Zool. Soc. Lon- 
don 1 (12): 133. Apr. 16, 1834 ("apud Ran- 
goon"; type locality here restricted to the hills 
of the Pegu District, Pegu Division, Burma). 

Eurylaimus lunatus Gould, Trans. Zool. Soc. Lon- 
don 1 (2): 176, pi. 25. Apr. 23-25, 1834 ("apud 
Rangoon, Peninsula? India? ulterioris"). 

Diagnosis. — The lores rusty or blackish 
rusty ; the sides of the head and the ear coverts 
ferruginous; the forehead pale ashy, this color 
changing insensibly to the light ferruginous of 
the crown and nape; the scapulars and upper 
back ashy brown; the lower back rufescent, 
this color changing to chestnut-rufous on the 
rump and upper tail coverts; the innermost 
secondaries rufous-buff. 

Range. — Pegu Yomas; Karenni; Tenasserim 
(south to the Amherst District); northwestern 
Siam (including Chiang Rai Province). 

Specimens examined. — Siam: Northwest: 
Chiang Saen Kao (1 male, 2 females), Wiang 
Khae (1 female), Ban Muang Sum (1 male), 
Doi Pha Horn Pok (1 male, 1 female), Doi Hua 
Mot (1 male, 2 females), Doi Chiang Dao 
(1 male, 2 females), Doi Suthep (1 male, 2 
females), Doi Khun Tan (1 male, 2 females); 
Tenasserim: Salween District: Papun (1 

Remarks. — The population I have taken to 
represent S. I. lunatus is immediately separable 
from all except the next following by the pale 
coloration of its innermost secondaries. 

It must be said, however, that no topo- 
typical material has been available to me, and 
the assumption that the bird of Pegu is the 
same as that of northwestern Siam rests only 
upon zoogeographical probability. The colored 
portrait of "lunatus" in Gould's Birds of Asia 
is, as observed by Hume (Stray Feathers 3: 
53. 1875), altogether too brightly colored; that 
given in the Trans. Zool. Soc. London could 
as easily picture the race described below from 
southwestern Siam. 

7. Serilophus lunatus intrepidus, n. subsp. 

Type— A.M.N.H. no. 203342, adult female, 
collected 2<S miles southeast of Ban Um Phang 
(a village at lat. 15°47' N., long. 98°50' E.), 
southwestern Siam, on February 2, 1924, by 
Arthur S. Vernay (original number 611). 

Diagnosis. — The lores rusty or blackish 
rusty; the sides of the head and the ear coverts 
ferruginous; the forehead pale ashy, this color 
changing insensibly to the light ferruginous of 
the crown and nape; the scapulars and upper 
back ashy brown; the lower back chestnut- 
rufous, this color changing to rufous-chestnut 
on the rump and upper tail coverts; the inner- 
most secondaries rufous-buff. 

Range. — Southwestern Siam and the adja- 
cent parts of Tenasserim (Amherst and Tavoy 

Specimens examined. — Tenasserim: Am- 
herst District: "Thaungyin valley" (1 male), 
"Thoungyin" (1 male), Moulmein (1 male), 
headwaters of the Mepale Chaung (1 female), 
Mitan (1 male); Siam: Southwest: 17 miles 
east of Lakya, Tenasserim (1 female), 28 miles 
southeast of Ban Um Phang (1 male, 1 female), 
20 miles west of Muang Kamphaeng Phet (1 

Remarks. — S. I. intrepidus is distinguished 
from all races except lunatus by the light- 
colored innermost secondaries. From lunatus 
itself it is separable only by the richer colora- 
tion of the posterior upper parts, in this show- 
ing normal intergradation between the more 
northern form and S. I. stolidus. 

It is probable that the differentiating charac- 
ter of the race would appear more strongly in 
specimens from more southerly localities; in 
default of such material, I have selected as type 
locality the most southern provenience repre- 
sented in my series. 

8. Serilophus lunatus stolidus Robinson 
and Boden Kloss 

Serilophus lunatus stolidus Robinson and Boden 
Kloss, Bull. Brit. Orn. Club 40: 16. Dec. 8, 1919 
(Khao Nong, a mountain in peninsular Siam at 
lat. 8°55' N., long. 99°38' E.). 

Diagnosis. — The lores ashy rusty; the sides 
of the head and the ear coverts ashy ferrugi- 
nous; the forehead and forecrown pale ashy, 
this color changing insensibly to the pale ashy 
ferruginous of the hindcrown and nape; the 

Mar. 15, 1948 deignan: the races of the silver- breasted broadbill 


scapulars and upper back ashy brown; the 
lower back chestnut-rufous, this color changing 
to rufous-chestnut on the rump and upper tail 
coverts; the innermost secondaries chestnut- 

Range. — Peninsular Siam (south of lat. 
11°40' N.); Tenasserim (Mergui District). 

Specimens examined. — Siam: Peninsula: 
Khao Luang, at lat. 11°40' N., long. 99°35' E. 
(1 female), Khao Luang, at lat. 8°30' N., long. 
99°45 / E. (3 males). 

Remarks. — Compared with "lunatus" from 
the Isthmus of Kra, Robinson and Boden 
Kloss found that stolidus had "deeper-coloured 
inner secondaries and tertiaries" and "slightly 
more drab, less fulvous ear coverts." One would 
not ordinarily expect isthmian examples (lat. 
10° N.) to differ in any important way from 
those of Khao Nong (lat. 8°55' N.), and the 
fact that the single specimen from lat. 1 1 °40 / 
N. is indistinguishable from those of lat. 8°30 / 
N. seems to show that only one race occurs in 
the Siamese portion of the Peninsula. 

9. Serilophus lunatus rothschildi Hartert 

and Butler 

Serilophus rothschildi Hartert and Butler, Bull. 
Brit. Orn. Club 7: 50. May 25, 1898 (Gunong 
Ijau, elev. 3,000 feet, Perak State, Malaya). 

Diagnosis. — The lores pale ashy; the sides 
of the head and the ear coverts ashy or brown- 
ish ashy; the forehead and forecrown pale 
ashy, this color changing insensibly to the 
brownish ashy or ashy brown of the hindcrown 
and nape; the scapulars and upper back deep 
ashy brown; the lower back chestnut-rufous, 
this color changing to rufous-chestnut on the 
rump and upper tail coverts; the innermost 
secondaries chestnut-rufous. 

Range. — Malaya (northern Perak to south- 
ern Selangor). 

Specimens examined. — Malaya: Perak 
State: Gunong Ijau (2 females);, Selangor 
State: Ginting Bidei (2 females). 

10. Serilophus lunatus moderatus Chasen 

Serilophus lunatus moderatus Chasen, Treubia 17 
(2): 137. July 1939 (Palok, near Mount Leuser, 
elev. ca. 1,000 meters, Achin, northern Su- 

Diagnosis. — The lores blackish ashy; the 
sides of the head and the ear coverts ashy or 
brownish ashy; the forehead and forecrown 
pale ashy, this color changing insensibly to the 
brownish ashy or ashy brown of the hindcrown 
and nape; the scapulars and upper back deep 
ashy brown; the lower back chestnut-rufous, 
this color changing to rufous-chestnut on the 
rump and upper tail coverts; the innermost 
secondaries chestnut-rufous. 

Range. — Northern Sumatra (Achin). 

Specimens examined. — Sumatra: North: 
Bandar Bahru (1 male), Blangnanga (2 males, 
2 females), Blangbeke (1 male, 1 female), 
Kungke (1 female). 

Remarks. — This form seems to be separable 
from rothschildi only b} r its darker lores. 

11. Serilophus lunatus intensus Robinson 
and Boden Kloss 

Serilophus lunatus intensus Robinson and Boden 
Kloss, Journ. Straits Branch Roy. Asiat. Soc. 
73 : 276. July 1916 (Siolak Dras, Korinchi Valley 
at elev. 3,100 feet, southwestern Sumatra). 

Diagnosis. — The lores blackish ashy; the 
sides of the head and the ear coverts brownish 
ashy; the forehead and forecrown pale ashy, 
this color changing insensibly to the brownish 
ashy or ashy brown of the hindcrown and nape; 
the scapulars and upper back deep ashy brown, 
lightly suffused with rufous; the lower back 
chestnut-rufous, this color changing to rufous- 
chestnut on the rump and upper tail coverts; 
the innermost secondaries chestnut-rufous. 

Range. — Sumatra (excepting Achin). 

Specimens examined. — Sumatra: Southwest: 
Siolak Dras (2 males, 1 female). 

Remarks. — Meyer de Schauensee and Ripley 
(Proc. Acad. Nat. Sci. Philadelphia 91: 336. 
1940) have combined moderatus with rothschildi 
but have kept intensus as a "thin" race, barely 
separable from their rothschildi-moderatus. All 
Sumatran birds seen by me, however, are 
easily distinguishable from true rothschildi by 
the color of the lores. To my. eyes, the real 
problem is to divide moderatus from intensus, 
and this seems to be just possible in series b} r 
the tone of the brown of the scapulars and 
upper back. 




Meeting of Executive Committee In view of the evident deficit the Committee 

turned its attention to ways and means of in- 
The Executive Committee met in the Cos- creasing the Academy income so that running 
mos Club at 8:10 p.m. on February 2, 1948, to expenses in the future might be met from yearly 
consider the budget submitted by the Treas- income without depleting the invested assets, 
urer. The President, F. D. Rossini, presided; Discuss i on was centered about two means: to- 
others present were: H. S. Rappleye, Walter creasing the numD er of paying members and 
Ramberu, W. L. Schmitt, and C. L. Gazin. initiating a drive for Journal subscriptions. 
After careful consideration of the separate About 18 vacancies exist in the Academy 

items, the Committee unanimously approved membership, and an informal discussion the 
the following budget for 1948: Committee showed agreement in favor of rais- 

ing the limit of membership to permit a greater 
Item Amount • £ ,, . 

income irom this source. 
Operating Expenses: With fd to the poss i b iii ty f obtaining 

IfecS C ° $ 400 additional funds from new subscriptions, the 

t™ o„™1 o^n Secretary was instructed to inform the Custod- 

S M & C. of Pubs. 50 $1 200 ian an0 - Subscription Manager of Publications 

that the Executive Committee is desirous of 

having the Subscription Manager attempt to 

Journal: secure these subscriptions through distribution 

Printing, mailing, illustra- f approximately 100 cooies of an appropriate 

tions, & reprints $4 , 776 issue of the j 0URNAL? e ach with a reprint of the 

Editorial Assistant 300 index of the JouRNAL for 1947 and a cover i n g 

Miscellaneous 50 5,126 , , . ._ , ,. n . ,. 

______ letter to various university and scientific msti- 

Total expenditures $6,326 tution libraries in this country and abroad not 

Estimated receipts 6 , 144 now subscribing to the Journal. 

The meeting was adjourned at 9:55 p.m. 

Deficit $ 182 C. L. Gazin, Secretary. 


Robert Hamilton Lombard was born in 
Ashburnham, Mass., on December 3, 1887, and 
died on October 11, 1947. He was graduated 
from Cushing Academy in Ashburnham in 
1906, and from Massachusetts Institute of 
Technology in 1910 with the S.B. degree. While 
at M.I.T. he was research assistant in physical 
chemistry. He published papers with Prof. M. 
DeKay Thompson and Prof. A. A. Noyes, un- 
der whom he worked. He obtained his Ph.D. 
degree at Columbia University in 1915 in 
chemistry and was an instructor there. He pub- 
lished two papers with Prof. Alexander Smith, 
one dealing with phosphorus pentachloride and 
the other with ammonium halides. 

From 1915 to 1927 Dr. Lombard was em- 
ployed as a chemist in the Geophysical Labo- 
ratory of the Carnegie Institution of Washing- 
ton, where he conducted research on the sys- 
tem copper-iron-sulphur. He published three 
papers on this work, with E. T. Allen and 
H. E. Merwin as co-authors. 

For the last 20 years Lombard was a research 
chemist in the Norton Company, Research 
Laboratories, at Worcester, Mass. He made 
important contributions to the abrasive in- 
dustry, and a number of patents, both domestic 

and foreign, have been granted covering his 
developments. Most of his researches were 
connected with the vitrified bonding of grind- 
ing wheels and other abrasive products. The 
most important new product resulting from his 
work is a grinding wheel having diamonds as 
the abrasive grains and a new-type vitrified 
bond. This product is superior to resinoid 
bonded and metal bonded diamond wheels in 
that the diamonds are held more firmly and 
efficiently while, at the same time, a freer cut- 
ting action is obtained. 

Dr. Lombard was a member of the honorary 
scientific societies Sigma Xi and Phi Lambda 
Upsilon. He was a member of the American 
Chemical Society, the Washington Academy 
of Sciences, the American Fern Society, and the 
Worcester Chemists' Club and was a fellow of 
the American Association for the Advancement 
of Science. His hobbies included botany and 
photography, and he had a collection of fine 
cameras. He was a member of the First Baptist 
Church of Worcester. 

He is survived by his wife, Hazel Soule Lom- 
bard, and a 12-year old daughter, Anne. 

L. H. Milligan. 

Officers of the Washington Academy of Sciences 

President Frederick D. Rossini, National Bureau of Standards 

Secretary C. Lewis Gazin, U. S. National Museum 

Treasurer Howard S. Rappleye, Coast and Geodetic Survey 

Archivist _ Nathan R. Smith, Plant Industry Station 

Custodian and Subscription Manager of Publications 

Harald A. Rehder, U. S. National Museum 

Vice-Presidents Representing the Affiliated Societies: 

Philosophical Society of Washington Walter Ramberg 

Anthropological Society of Washington T. Dale Stewart 

Biological Society of Washington John W. Aldrich 

Chemical Society of Washington Charles E. White 

Entomological Society of Washington C. F. W. Muesebeck 

National Geographic Society Alexander Wetmore 

Geological Society of Washington William W. Rubey 

Medical Society of the District of Columbia Frederick O. Coe 

Columbia Historical Society. . . Gilbert Grosvenor 

Botanical Society of Washington Ronald Bamford 

Washington Section, Society of American Foresters William A. Dayton 

Washington Society of Engineers Clifford A. Betts 

Washington Section, American Institute of Electrical Engineers 

Francis B. Silsbee 

Washington Section, American Society of Mechanical Engineers 

Martin A. Mason 

Helminthological Society of Washington Aurel O. Foster 

Washington Branch, Society of American Bacteriologists Lore A. Rogers 

Washington Post, Society of American Military Engineers . Clement L. Garner 

Washington Section, Institute of Radio Engineers Herbert Grove Dorsey 

Washington Section, American Society of Civil Engineers Owen B. French 

Elected Members of the Board of Managers: 

To January 1949 Max A. McCall, Waldo L. Schmitt 

To January 1950 F. G. Brick wedde, William W. Diehl 

To January 1951 Francis M. Defandorf, William N. Fenton 

Board of Managers All the above officers plus the Senior Editor 

Board of Editors and Associate Editors [See front cover] 

Executive Committee Frederick D. Rossini (chairman), Walter Ramberg, 

Waldo L. Schmitt, Howard S. Rappleye, C. Lewis Gazin 

Committee on Membership 

Harold E. McComb (chairman), Lewis W. Butz, C. Wythe Cooke, William 

W. Diehl, Lloyd D. Felton, Regina Flannery, George G. Manov 

Committee on Meetings Raymond J. Seeger (chairman), 

Frank P. Cullinan, Fred L. Mohler, Francis O. Rice, Frank Thone 

Committee on Monographs: 

To January 1949 Lewis V. Judson (chairman), Edward A. Chapin 

To January 1950 Roland W. Brown, Harald A. Rehder 

To January 1951 William N. Fenton, Emmett W. Price 

Committee on Awards for Scientific Achievement (Karl F. Herzfeld, general chairman): 

For the Biological Sciences 

C. F. W. Muesebeck (chairman), Harry S. Ber.vton, Chester W. Emmons, 
Elmer Higgins, Mario Mollari, Gotthold Steiner, L. Edwin Yocum 

For the Engineering Sciences 

Harry Diamond (chairman), Lloyd V. Berkner, Robert C. Duncan, 
Herbert N. Eaton, Arno C. Fieldner, Frank B. Scheetz, W. D. Sutcliffe 

For the Physical Sciences 

Karl F. Herzfeld (chairman), Nathan L. Drake, Lloyd D. Felton, 
Herbert Insley, William J. Rooney, Robert Simha, Michael X. Sullivan 

Committee on Grants-in-aid for Research 

. .F. H. H. Roberts, Jr. (chairman), Anna E. Jenkins, J. Leon Shereshevsky 

Representative on Council of A. A. A. S Frank Thone 

Committee of Auditors 

William G. Brombacher (chairman), Harold F. Stimson, Herbert L. Haller 

Committee of Tellers 

...John W. McBurney (chairman), Roger G. Bates, William A. Wildhack 



Philology. — English-language surnames of biological origin. Perez 

Simmons 81 

Botany. — New species of trees from western Ecuador. Elbert L. 

Little, Jr 87 

Herpetology. — A collection of salamanders from Mount Rogers, 

Virginia. Richard L. Hoffman and Hubert I. Kleinpeter 106 

Ornithology. — The races of the silver-breasted broadbill, Serilophu6 

lunatus (Gould). H. G. Deignan 108 

Proceedings: The Academy 112 

Obituary: Robert Hamilton Lombard 112 

This Journal is Indexed in the International Index to Periodicals 

J K 

Vol. 38 April 15, 1948 No. 4 




James I. Hoffman Alan Stone Frank C. Kracek 



Lawrence A. Wood Richard E. Blackwelder 


J. P. E. Morrison James S. Williams 


Elbert L. Little, Jr. Waldo R. Wedel 


Irl G. Schoonover 


published monthly 



450 Ahnaip St. 

at Menasha, Wisconsin 


Entered as second class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a special rate of postage provided for in the Act of February 28, 1925. 

Authorised January 21, 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof, — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alteration! 
made in the proof by the author, may be charged to the author. 

Author' 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate) : 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 .50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8.00 .60 

Vol. 23 to current vol. (12 numbers per vol.) 7 . 50 . 90 

• Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 




Vol. 38 

April 15, 1948 

No. 4 

BIOCHEMISTRY. — The chemical nature of enzymes. 1 James Batcheller Sum- 
ner, Cornell University. (Communicated by Waldo L. Schmitt.) 

Why was it difficult to isolate an enzyme? 
Here I employ the word "isolate" as mean- 
ing preparing in pure condition. It does not 
seem difficult to isolate and crystallize an 
enzyme now, but it was difficult 20 years 
ago. The reasons were several. One was the 
inertia of men's ideas. Another reason was 
the influence exerted by Willstatter and 
his school, who held that enzymes were 
neither lipids, carbohydrates, nor proteins 
and who believed enzymes to exist in ex- 
cessively low concentrations in plants and 
animals. The misconception that colloidal 
substances had a chemistry different from 
that of crystalloids was another stumbling 
block. A yet more important reason why 
enzymes could not be isolated readily was 
because work with these substances requires 
a special technique. The enzyme is present 
in relatively low concentration in the raw 
material. It often is highly unstable. The 
raw material nearly always contains natural 
protectors which are left behind during the 
purification processes. In the absence of 
these natural protectors, the enzyme may 
become destroyed by traces of heavy metals, 
by oxidation, by unfavorable pH, or even 
by autolysis. 

I wish to tell next why I decided in 1917 
to attempt to isolate an enzyme. At that 
time I had little time for research, not much 
apparatus, research money, or assistance. 
I desired to accomplish something of real 
importance. In other words, I decided to 
take a "long shot." A number of persons 
advised me that my attempt to isolate an 

1 Nobel Laureate Lecture delivered at Stock- 
holm, Sweden, December 12, 1946. Previously- 
published by the Royal Academy of Sciences, 

enzyme was foolish, but this advice made 
me feel all the more certain that if success- 
ful the quest would be worth while. 

The reasons why I chose to work with 
urease were several. I had been working 
with urease as a reagent for the estimation 
of urea in muscle, blood, and urine. This 
urease was prepared from soy bean meal. 
In 1916, Mateer and Marshall found that 
the jack bean, Canavalia ensiformis, con- 
tained about 16-fold more urease than the 
soy bean. The jack bean appeared to me 
to be extraordinarily rich in urease and I 
could see no reason why this enzyme could 
not be isolated in pure form and charac- 
terized chemically. Claude Bernard has said 
that success or failure may depend upon 
the lucky choice of some reagent or raw 
material. Willstatter was unfortunate in 
his choice of saccharase as an enzyme to 
isolate. I was fortunate in choosing urease. 

I hoped that urease would turn out to be 
a globulin, since globulins usually, if not 
always, are present in beans and seeds, and 
since globulins are easy to precipitate by 
dialysis. Other reasons for choosing urease 
as the enzyme to isolate were that this 
enzyme can be estimated quantitatively 
very accurately and readily, and that 
urease could be expected to be one enzyme, 
rather than a mixture of enzymes, acting 
as it does on such a simple substrate as urea. 

I started trying to isolate urease in the 
fall of 1917, having been occupied previ- 
ously with analytical methods. At this 
time, our laboratory contained no adequate 
apparatus for grinding jack beans. We first 
used a coffee mill and then ground the 
coarse material with a mortar and pestle. 
Years later, we constructed a mill which 
was run by an electric motor and which 





employed a bolting device, However, in the 
meanwhile we usually used commercial 
jack bean meal. This commercial meal was 
not always satisfactory. 

The jack bean is a miniature world in 
itself and contains all of the elements re- 
quired for life, growth, and reproduction. I 
decided to isolate and characterize as many 
as possible of the chemical compounds pres- 
ent in the bean. I found various minerals, 
proteins, carbohydrates, lipids, extractives, 
pigments, and enzymes to be present. Par- 
ticular attention was paid to the proteins 
of the jack bean, since I expected to find 
that urease was one of these. I isolated two 
globulins in crystalline form and named 
these concanavalin A and concanavalin B. 
A third globulin which I called "canavilin" 
separated as spheroids upon dialysis. Years 
later Howell and I were able to crystallize 
this, after a preliminary digestion with 
trypsin. Many years later we discovered 
that concanavalin A is a hemagglutinin for 
the red cells of certain animal species as 
well as an excellent precipitant for certain 
polysaccharides. A lipid fraction of the 
jack bean was observed to function as a 
thromboplastic agent. 

In attempting to concentrate and purify 
urease, I employed fractional precipitation 
with alcohol, acetone, and other organic 
solvents. Fractional precipitation with am- 
monium sulfate, magnesium sulfate, and 
other neutral salts was tried. I tested a large 
number of salts of heavy metals or precipi- 
tants. I employed a very large number of 
reagents as absorbents. This work covered 
many years. At times I grew discouraged 
and temporarily abandoned the quest, but 
always returned to it again. 

At first we used to extract urease from 
jack bean meal with water. These aqueous 
extracts were viscous and therefore very 
difficult to filter. Glycerol extracts w^ere 
even more bothersome. I learned that 
Folin used 30 percent alcoholic extracts of 
jack bean meal as a source of urease for 
analytical purposes. It was found that ex- 
traction with 30 percent alcohol was of 
distinct advantage, inasmuch as this sol- 
vent dissolved most of the urease but 
failed to dissolve a rather large quantity of 
the other proteins. Hence, a considerable 

purification was achieved through the use 
of this solvent. The alcoholic extracts 
filtered very rapidly, leaving the undis- 
solved material behind on the filter paper. 
The only disadvantage of 30 percent alco- 
hol lay in the slow inactivating action of 
this solvent upon the urease. However, if 
kept at low temperatures, there was no 
inactivation of the enzyme. 

When kept at low temperatures, 30 per- 
cent alcoholic extracts of jack bean meal 
formed precipitates. These precipitates con- 
tained practically all of the urease, together 
with concanavalin A, concanavalin B, and 
other proteins. At this time we had no ice 
chest in our laboratory and we used to place 
cylinders of 30 percent alcoholic extracts on 
our window ledges and pray for cold 

It seemed to me of interest to employ 
dilute acetone instead of 30 percent alcohol 
and to see whether this substitution would 
result in any improvement in the method of 
purification. Accordingly I diluted 316 cc 
of pure acetone to 1,000 cc and used this as 
the means of extracting the urease. I rou- 
tinely employed this dilution of acetone, 
since I had been preparing 30 percent alco- 
hol through diluting 95 percent alcohol in 
this manner. The acetone extract was 
chilled in our newly acquired ice chest over- 
night. The next morning I examined the 
filtrate. It contained practically no precipi- 
tate, thus differing from alcoholic filtrates. 
However, upon observing a drop of the 
liquid under the microscope, it was seen to 
contain many tiny crystals. These were of a 
shape that I had never observed previously. 
I centrifuged off some of the crystals and 
observed that they dissolved readily in 
water. I then tested this water solution. 
It gave tests for protein and possessed a 
very high urease activity. I then telephoned 
to my wife, "I have crystallized the first 

Now I should like to tell this audience 
what enzyme crystals look like. This de- 
scription applies also to proteins, since 
enzymes are proteins. Enzyme crystals are 
nearly always microscopic, since these 
compounds, being of high molecular weight, 
diffuse relatively slowly and therefore 
crystallize slowly. Enzyme crystals belong 

Ape. 15, 1948 



nearly always to the isometric or hexagonal 
systems. Sometimes enzymes separate from 
solution as spheroids. This formation indi- 
cates a tendency to crystallize, as Dr. 
Northrop can tell you from his experience. 
Spheroids are nearer the crystalline state 
than purely amorphous material, such, for 
example, as casein which has been precipi- 
tated from milk by the addition of acetic 
acid. Spheroids have sometimes been found 
by us to be aggregations of many needles 
either in parallel or concentrically arranged. 

Having convinced myself that I really 
had isolated urease in crystalline form, I 
read a paper on this matter at Clifton 
Springs, N. Y., and published an article 
in the August number of the Journal of 
Biological Chemistry for the year 1926. But 
now a difficulty arose. The commercial jack 
bean meal which we had been using suffered 
a decline in quality and we could obtain no 
urease crystals from it unless we added a 
small amount of acetic acid to the alcoholic 
filtrates. Even then, the yield of crystals 
was low. Analyses showed that the recent 
samples of jack bean meal contained only 
about one-half as much urease as the ear- 
lier samples. Accordingly, attempts were 
made to obtain satisfactory meal, or satis- 
factory jack beans. We grew jack beans in 
one of the Cornell greenhouses. The beans 
grew poorly and the yield was less than the 
number of beans planted. We obtained 
jack beans from Texas, Guatemala, and 
Honduras, but these were low in urease. I 
happened to meet a plant physiologist, Dr. 
Albert Muller, who said he would grow me 
some jack beans at Mayaguez, Puerto Rico. 
I gave him about a kilo of jack beans rich in 
urease, the last I had left. Some seven 
months later about a bushel of beans ar- 
rived from Puerto Rico. These beans were 
rich in urease. The finely ground meal gave 
a high yield of urease crystals. Later we 
were able to obtain excellent jack beans 
from an Arkansas farmer. He has supplied 
us with jack beans ever since. 

I wish to speak now about proofs of the 
identity of the crystals with urease. In 
cases of this sort, one piece of experimental 
evidence is not sufficient to constitute a 
valid and satisfactory proof; one must sub- 
mit many pieces of evidence. At this time 

I had no access to the ultracentrifuge of 
Svedberg nor to the electrophoresis appara- 
tus of Tiselius. However, I was able to offer 
evidence of another sort, as is shown below : 

1. When the crystals separated there 
occurred a great increase in purity, namely 
from 700 to 1,400-fold. Using other methods 
the increase in purity observed was very 
much less than this and at times there even 
occurred a decrease in purity. 

2. When the distribution of urease was 
followed quantitatively it was found that 
of the urease passing into the filtrate as 
much as 40 percent or more separated with 
the crystals. 

3. Recrystallization increased the purity 
of the urease. 

4. Adding traces of poisons, such as 
mercuric chloride or formaldehyde, to jack 
bean meal inactivated the urease and, at 
the same time, prevented the appearance of 
the crystals. 

In obtaining the crystals, I felt much the 
same as a person does who is trying vainly 
to place in position a piece of a machine. 
Suddenly the piece slides in as if covered 
with butter. One knows that it is now where 
it belongs. 

During later work on crystalline urease, 
I was fortunate to have a number of excel- 
lent men in our laboratory. These were 
Doctors Hand, Kirk, Poland, and Dounce. 
We found that trypsin neither digested 
nor inactivated urease. Since trypsin is a 
proteolytic enzyme of second attack, it 
does not digest certain native proteins 
readily. Some proteins that are digested 
with great difficulty are hemoglobin, oval- 
bumin, and the serum proteins. After urease 
had been denatured by acid or by heating, 
it was found to be very readily digested by 

Pepsin acts best in a strongly acid me- 
dium and strong acid rapidly destroys ure- 
ase. However, at pH 4.3 we found that 
urease was so slowly destroyed that it was 
possible to demonstrate a parallel digestion 
and inactivation by pepsin. In place of 
pepsin it was possible to use papaincysteine. 

In our laboratory Dr. J. Stanley Kirk 
was able to immunize rabbits to crystalline 
urease. Kirk started by giving rabbits as 
little as 0.03 mg of crystalline urease intra- 



peritoneally. This dose was given twice 
weekly and was finally increased to 1 ,000 
lethal doses. The immunized rabbits con- 
tained antiurease in their blood serum. This 
antiurease could be purified by precipitat- 
ing it by adding urease, washing the urease- 
antiurease precipitate, decomposing the 
complex with 0.05 N hydrochloric acid, 
bringing to pH 5.0, and centrifuging down 
the denatured urease. The antiurease was 
not harmed by this treatment and could 
be employed as an excellent precipitant in 
testing for urease. It gave a visible precipi- 
tate with solutions of urease diluted 1 to 
500,000. However, urease allowed to stand 
for a few seconds with 0.05 N hydrochloric 
acid and then neutralized gave no precipi- 
tate with antiurease; neither did it possess 
any urease activity. 

Northrop has made good use of physical 
methods to demonstrate the identity of 
pepsin with his pepsin crystals. While we 
have not employed such methods, Kubo- 
witz and Haas, working in Warburg's labora- 
tory, have demonstrated that ultraviolet 
light is absorbed by highly dilute solutions 
of crystalline urease and that exactly the 
same wavelengths that are absorbed are 
those which destroy urease. 

In 1930, Northrop of the Rockefeller 
Institute obtained pepsin in crystalline 
form. A short time later, Northrop and 
Kunitz obtained crystalline trypsin, crys- 
talline chymotrypsin, and also the zymo- 
gens of these enzymes in crystalline form. 
Their monumental work was of very great 
help in bringing the scientific world to ad- 
mit that enzymes can be isolated in pure 
and crystalline condition. In this connec- 
tion I wish to note that Professor von Euler 
aided me greatly when I worked on urease 
in his laboratory at Stockholm's Hogskule 
in 1929 and that I received valuable help 
in 1937 while working in the laboratory of 
Prof. The Svedberg at the University of 

The announcement of the crystallization 
of urease and pepsin was not accepted by 
some biological chemists. In Germany stu- 
dents of Willstatter attempted to show that 
our crystalline proteins were merely carriers 
of the enzymes. It suffices to say that these 
attempts to disprove our work failed, as 

they were bound to fail, since Northrop and 
I were right and since our evidence was 

To date about thirty enzymes have been 
obtained in crystalline and in presumably 
pure condition. Certain enzymes are col- 
ored, being combinations of specific pro- 
teins with such prosthetic groups as ribo- 
flavin, phosphate, or hematin. Theorell 
has crystallized the yellow enzyme of War- 
burg and Christian and, remarkably enough, 
has been able to split the protein from the 
riboflavin phosphate and later to reunite 
these two components. He has shown con- 
clusively that the union of these compo- 
nents proceeds stoichiometrically. 

The enzymes catalase and peroxidase 
are compounds of protein with hematin and 
these two enzymes possess characteristic 
absorption bands. This property has greatly 
facilitated the crystallization of these two 
enzymes. In addition, it should be noted 
that these two enzymes have been shown 
to form compounds with hydrogen sulfide, 
sodium fluoride, sodium azide, etc. 

All enzymes are proteins but not all pro- 
teins are enzymes. Many, if not all enzymes 
can be crystallized. The oxidizing enzymes 
all appear to be conjugated, or compound, 
proteins, while the hydrolytic enzymes are, 
as far as we can tell, without prosthetic 
groups. However, even hydrolytic enzymes 
must have reactive groups. Every enzyme 
requires a specific method for its purifica- 
tion. Our present methods of purifying 
both enzymes and proteins are crude and 
unsatisfactory for the most part. It is fairly 
certain that better methods wall be dis- 
covered in the near future. One can purify 
an enzyme either by precipitating the im- 
purities or else by precipitating the enzyme. 
The latter procedure is to be preferred, but 
the former procedure often is necessary at 
the beginning. When I speak of precipitat- 
ing the enzyme I mean, of course, a more or 
less specific precipitation and not a general 
precipitation of everything, much as occurs 
on adding a great excess of alcohol or ace- 

Some day every enzyme in living matter 
will have been discovered and described. 
Every chemical reaction which goes on will 
have been recorded. We probably can ex- 

Apr. 15, 1948 

beebe: brunauer's contributions in adsorption 


pect to find some enzymes which are glyco- 
proteins, others which are lipoproteins, and 
others which are nucleoproteins. 

In 1917, the role played by enzymes was 
only partially understood. Digestive en- 
zymes were well known, autolytic and oxi- 
dative enzymes were somewhat known, but 
not well understood. At present we realize 
the tremendous complexity of the cell. In 
muscle alone, some 60 enzymes are known 
to occur. Thanks to relatively recent in- 
vestigation, practically all of the compli- 
cated reactions involved in the breaking 
down of glycogen to carbon dioxide and 
Avater have been made clear. 

The organic chemist has never been able 
to synthesize cane sugar, but, by using en- 
zymes, the biological chemist can synthesize 
not only cane sugar, but also gum dextran, 
gum levan, starch, and glycogen. 

We know now of the existence of enzymes 
which employ phosphoric acid instead of 
water and which might be given the general 
term "phosphorases." There are the phos- 
phorylases, transphosphorylases, phospho- 

isomerases, phosphomutases, and phospho- 

From the work of Cori and his associates, 
we now have evidence that hormones func- 
tion through their effect on enzymes. Thus, 
glucose is transformed into glucoses-phos- 
phate when it reacts with ATP in the pres- 
ence of the enzyme known as hexokinase. 
This reaction, essential for utilization of 
glucose, is inhibited by the diabetogenic 
hormone coming from the anterior pitui- 
tary. This inhibiting action is abolished by 

We can sum up by saying that, as the re- 
sult of discoveries in the field of enzyme 
chemistry, some questions have been an- 
swered and many new questions have 
arisen. We live in an expanding universe in 
more senses than that of the astronomers. 

In conclusion, I wish to pay tribute to my 
former teacher, Otto Folin, who emigrated 
as a boy of 17 from Sm aland to America and 
who, as professor of biochemistry at Har- 
vard, inspired me as he did many other 
young men. 

CHEMISTRY. — Dr. Stephen Brunauer's contributions in the field of adsorption. 1 
Ralph A. Beebe, Amherst College. (Communicated by James I. Hoffman.) 

At the outset I think it is desirable to 
distinguish clearly between physical ad- 
sorption and chemisorption, and with that 
in view I have tabulated below a number of 
characteristics of each type of process: 

Physical Adsorption Chemisorption 

1. Rapid Rapid or slow 

2. Reversible Often irreversible 

3. Low heat High heat 

4. Non-specific Specific 

5. Multilayer Monolayer only 

Since this comparison should perhaps be 
discussed in some more detail, attention is 
called to the following : 

1. Physical adsorption is generally a 
rapid process whenever the surface is read- 
ily accessible to the gas or vapor being ad- 
sorbed, although one may encounter a slow 

1 Address delivered before the Chemical Society 
of Washington, March 1946, on the occasion of 
the awarding of the Society's Hillebrand Prize to 
Dr. Stephen Brunauer. Dr. Beebe's preliminary 
congratulatory remarks are here omitted. Re- 
ceived September 24, 1947. 

adsorption when it is necessary for the ad- 
sorbate to diffuse into cracks or capillaries 
to reach the adsorbing surface. Chemisorp- 
tion, on the other hand, like chemical reac- 
tions in general, may be rapid or slow, and 
frequently displays a temperature coeffi- 
cient requiring an activation energy of the 
order of magnitude found in chemical reac- 

2. Physical adsorption is in general re- 
versible, and it is possible to remove the 
adsorbate from the surface by outgassing 
at the temperature of adsorption or at 
slightly elevated temperatures. Chemisorp- 
tion, on the other hand, may be truly ir- 
reversible, indeed to such an extent that 
the adsorbate can be removed only by 
chemical action at elevated temperatures. 
For instance, oxygen chemisorbed on tung- 
sten can be removed only by heating to 
about 2000°C. Other chemisorbed layers, 
however, can be removed by less drastic 
treatment, let us say by outgassing at 



temperatures of a few hundred degrees 
above that at which the adsorption process 
takes place. 

3. In general, the heat of adsorption 
which indicates the energy of binding of the 
adsorbate molecules to the solid surface is 
of a low order of magnitude in the case of 
physical adsorption, usually two to three 
times the heat of vaporization for the first 
portions of adsorbate added to the surface. 
This means that for such gases as nitrogen, 
oxygen, argon, or carbon monoxide, all of 
which boil at temperatures approximately 
that of liquid air, the heat of adsorption 
will not exceed 5 kcals per mole. With 
larger molecules, such as butane, we en- 
counter higher heats of physical adsorption, 
which, however, do not exceed 15 to 16 
kcals per mole. Again one finds a marked 
difference in chemisorption, the heats of 
this process, like those of ordinary chemical 
reactions, running as high as 100 kcals per 
mole in certain instances. 

4. Although, as is pointed out in Dr. 
Brunauer's book (1), there has never been 
given an exact definition of the term spe- 
cificity as applied to adsorption, yet this 
term is of some use in a qualitative sense. 
It may be said that in the case of physical 
adsorption the heats evolved tend to be 
nonspecific to the chemical nature of the 
gases or vapors involved if these adsorbates 
have molecular weights, and therefore heats 
of vaporization, of the same order of mag- 
nitude For instance, one finds that the four 
substances mentioned above have heats of 
adsorption of the same order of magnitude. 
However, we shall see in a later portion of 
this paper that we must not generalize too 
far about this nonspecificity of physical 
adsorption, and that, indeed, significant 
differences become apparent especially 
when we study the physical adsorption of 
the same adsorbate on different surfaces. As 
might be expected, the heat of adsorption 
in chemical processes is highly specific and 
may vary all the way from, let us say, 15 
kcals to 100 kcals in different instances. In 
this respect, of course, chemisorption again 
resembles chemical reactions in general. 

5. It has been well established by the 
work of Dr. Brunauer and others that the 
forces which give rise to chemisorption ex- 

tend to a negligible extent beyond the 
monolayer of adsorbed molecules. On the 
other hand the work of the group with 
which Dr. Brunauer has been associated 
seems to establish beyond reasonable 
doubt that we are concerned in physical 
adsorption with the formation of multi- 
layers of adsorbate on the solid surface (2). 

With this background, we are now pre- 
pared to turn to a discussion and review of 
of some of the significant work to which Dr. 
Brunauer has contributed. In making a 
selection from Dr. Brunauer's experiments, 
I have chosen the work that has been es- 
pecially helpful to my own investigations. 

In Fig. 1 are shown the isotherms ob- 
tained by Emmett and Brunauer (3) for the 
adsorption of carbon monoxide on two iron 
catalysts at the temperature of liquid oxy- 
gen. Looking first at the isotherms for the 
pure iron catalyst, it is evident that the 
curve labeled "total adsorption" lies very 
much above the "physical adsorption" 
curve. For reasons which have been abun- 
dantly established by the group with which 
Dr. Brunauer has been associated, the 
points labeled B on each curve represent 
approximately the completion of the sta- 
tistical monolayer of physically adsorbed 
gas. Thus the volume indicated by Point B 
in the "physical adsorption" curve is the 
volume of gas required to complete a layer 
of carbon monoxide molecules in direct con- 
tact with the solid surface. The curve la- 
beled "total adsorption" was obtained by 
outgassing the sample at 400°C and then 
admitting carbon monoxide to the system 
at — 183°C. After this experiment, the sur- 
face was outgassed at —78°, or indeed 
within the range from —78° to 0°, the tem- 
perature of outgassing not being very criti- 
cal within this range. Following the out- 
gassing process, the adsorption of carbon 
monoxide was again measured and was 
found to follow the curve labeled "physical 
adsorption." Emmett and Brunauer inter- 
preted these experiments to indicate that in 
the initial experiment the carbon monoxide 
was adsorbed first as a chemisorbed layer 
with the carbon monoxide molecules held 
by relatively strong chemical bonds to the 
underlying atoms, and next as a physically 
adsorbed layer on top of the chemisorbed 

Apr. 15, 1948 beebe: brunauer's contributions in adsorption 


molecules. On outgassing in the tempera- 
ture range from — 78° to 0°C, the physically 
adsorbed molecules were removed, but the 
thermal energy was insufficient for the re- 
moval of the chemisorbed layer, with the 
result that the extent of physical adsorption 
could be determined by a subsequent ad- 
sorption measurement. It is obvious then 
that the difference in heights of the two 
curves represents chemisorption. 

From the isotherms for the promoted iron 
catalyst given in Fig. 1, it is evident again 
that the two types of adsorption process 
can be separated. In this case, the relatively 
large physical adsorption appears to indi- 
cate that the small percentage of promoter 
material present in the catalyst must cover 
a relatively large fraction of the surface 
since it is assumed that carbon monoxide is 
chemisorbed only on that fraction of the 
surface in which the iron atoms are exposed, 
but is physically adsorbed on the whole sur- 
face. This deduction comprises an impor- 
tant contribution by Emmett and Brunau- 
er to the understanding of promoter action. 

Following the completion of these experi- 
ments, Dr. Emmett suggested to me that it 
would be worth while to apply to this prob- 
lem the technique developed at Amherst 
for the calorimetric measurement of heats 
of adsorption. In Fig. 2 are shown the re- 
sults of these calorimetric experiments (4). 
Emmett and Brunauer, from an interpreta- 
tion of their curves, would have predicted 
that the chemisorption process would be 
superseded by physical adsorption after 
0.24 cc and 0.21 cc of carbon monoxide at 
— 78 and —183° respectively had been ad- 
sorbed per gram of catalyst. The initial 
high values in the differential heats of ad- 
sorption, of course, indicate chemisorption 
at 0°, -78°, and even at -183°, although 
the heat values are lower at the latter tem- 
perature. The sharp drop in the heats of 
adsorption at — 78° to the region of 4 kcals 
per mole at 0.25 cc per g, the point pre- 
dicted by Emmett and Brunauer as a result 
of the analysis of their isotherms, supplies 
experimental confirmation of the validity 
of their interpretations. This confirmation 
from the heat data at — 183°, while not so 
strikingly apparent, is nevertheless seen to 
be present. 

Next we shall turn our attention to an- 
other important facet of the work to which 
Dr. Brunauer has contributed heavily. I 
refer to the experimental and theoretical 
development of the theory of multimolecu- 
lar adsorption, known as the Brunauer, 
Emmett, and Teller theory {2) . In the early 
development of the experimental work, 
Emmett and Brunauer came to the tenta- 
tive conclusion that the monolayer is com- 
plete at the Point B of the characteristic 
isotherms obtained in many adsorption sys- 
tems and illustrated in Fig. 1. This conclu- 
sion w T as found to be consistent with the 
theoretical considerations developed by 
Brunauer, Emmett, and Teller. The funda- 
mental equation most generally used is 
given below: 

p/v(p -p) 



In this equation, v represents volume of 
vapor adsorbed, p the equilibrium pressure, 

Fig. 1. — Isotherms for CO on iron catalysts at 
— 183°: Pure iron catalyst, 973; promoted iron 
catalyst (K,0, 1.59 percent, and A1 2 3 , 1.3 per- 
cent), 931. 



and p the saturation pressure of the vapor 
at the temperature of the experiment, v m 

the volume of gas necessary to fill the mono- 
layer, and C a constant which is approxi- 
mately defined as C = e {El ~ EL),RT in which 
Ei is the heat of adsorption in the first 
layer, and El is the heat of vaporization, 
and R and T have their usual meaning, 
E h El, and R being expressed in calories 
and T as the absolute temperature. It is 
apparent that a straight line should be ob- 
tained if the function (p/v(po — p)) is plotted 
against p/po, and indeed literally scores if 
not hundreds of cases provide experimental 
data which conform to this condition. From 
such a plot, known as the B.E.T. plot, it is 
possible to obtain the slope and intercept 
of the straight line, and from these to calcu- 
late the values of v m and of C. Knowing v m 
and making an estimate of the area occu- 
pied by the adsorbate molecule on the sur- 
face, we can calculate the surface area. 
Obviously such an estimate of surface area, 
especially on finely divided or porous solids, 
is a welcome additional technique in the 
field of adsorption, and as a result has come 
into very wide use. It will become apparent, 



v. - 



b^— Ol_?_____q_ 

I think, in the experiments fco be described 
below, that our ability to estimate the sur- 
face area has added greatly to the signifi- 
cance of the interpretation of the experi- 
mental results. 

Now I want to tell you about some work 
we have been doing in Amherst on the heats 
of adsorption of vapors on carbon black 
(5). This work was initiated about a year 
ago as a result of a series of conversations 
between the writer and Dr. Walter R. 
Smith of Godfrey L. Cabot, Inc. At that 
time Dr. Smith was especially interested in 
getting at the fundamental reasons for the 
effectiveness of carbon blacks in reinforcing 
rubber. The investigations in his laboratory 
had indicated that different carbon blacks 
presented a very high degree of specificity 
in their reinforcing ability, especially as de- 
termined by the modulus of elasticity. It is 
apparent, since the carbon black is dis- 
persed through the rubber as particles of 
the order of several hundred Angstroms di- 
ameter, that the binding forces between the 
particles of black and the rubber must be 
adsorption forces between the surface of 
the blacks and the adjacent sections of the 
rubber molecules. FolloAving this line of 
reasoning, one would be tempted to predict 
that a close correlation might be found be- 
tween the extent of surface of a carbon 
black pigment and its reinforcing ability. 

That the above prediction is not borne 
out by the experimental facts is shown in 
Table 1, which contains data on two sam- 
ples selected from the several blacks which 
have been studied; these are labeled Spher- 
on Grade 6 and Graphon respectively. More 

Table 1 




300% eloDg. 

Spheron Grade 6 



sq. m./g. 

lb. sq. in. 

03 O.b 0<? 

V«( a<Js., cc /$• 

Fig. 2. — Heats of adsorption of CO on catalyst 
931: -183°, O, •; -78°, 0; 0°, Q. 

complete data on these blacks especially as 
regards the chemical composition of their 
surfaces and their underlying crystal struc- 
ture have been published elsewhere (5). 
Spheron Grade 6 is a channel black having 
good rubber reinforcing ability. The Graph- 
on has been prepared from the Spheron 

Apr. 15, 1948 beebe: brunauer's contributions in adsorption 








^r\— n 



Fig. 3. — Heats of adsorption on carbon blacks 
at -195°C: N 2 on Spheron Grade 6, O; N 2 on 
Graphon # 

Grade 6 by electrical heating to tempera- 
tures of 2,800 to 3,000°C (6). As is seen in 
Table 1, this heating process had relatively 
little effect on the particle diameters as de- 
termined by the electron microscope and on 
the specific surface areas as determined by 
the B.E.T. nitrogen adsorption method, al- 
though there was a profound effect on the 
rubber reinforcing ability as indicated by 
the modulus of rubber samples reinforced 
by the two pigments. 

Although there appeared to be no definite 
correlation between the extent of surface 
and the rubber reinforcing ability, it seemed 
possible that there might be a significant 
difference in state of surface if the two 
blacks were compared. Since a convenient 
means of testing the quality or state of sur- 
face of these carbon black samples seemed 
to be offered by data on heats of adsorption, 
calorimetric measurements have been un- 
dertaken using the elementary gases nitro- 
gen, oxygen, and argon at — 195°C (7). The 
results of the experiments with nitrogen are 
shown in Fig. 3 in which the differential 
heats of adsorption are plotted against the 

number of layers adsorbed (v/v m = 1 repre- 
sents a monolayer). 

In relation to the rubber reinforcing prob- 
lem, the difference in behavior of the Graph- 
on and the Grade 6 black is of great interest. 
It is apparent that although the high tem- 
perature treatment, by which the Graphon 
was made from the Grade 6 carbon black, 
resulted in a relatively small change in the 
total surface of the black, this treatment 
nevertheless produced a profound change in 
the energy state of the surface, with the re- 
sult that approximately the first third of the 
surface would hold adsorbate molecules 
considerably less firmly than the original 
Spheron Grade 6. Thus there appears to be 
a correlation between the rubber reinforcing 
abilities of the two blacks and the energy 
states of their surfaces as indicated by the 
heats of adsorption. 

Apart from the problem of rubber rein- 
forcement several other conclusions of con- 
siderable general interest may be drawn 
from the data of Fig. 3. (1) The order of 
magnitude of the heats indicates that the 
process is exclusively physical adsorption. 
(2) The rapid decrease in the differential 




J9 ^ 

©© © 

J •© 9 { 







Fig. 4. — Heats of adsorption of Spheron Grade 
6 carbon black at 0°C: n-butane, O; 1-butene, #; 
cis-2-butene, Q; trans-2-butene, ©. 



heats of adsorption for the first pails of the 
surface covered on the Grade 6 carbon 
black suggests a considerable heterogeneity 
of the surface with the fractions first cov- 
ered exhibiting excessively high binding 
energies. (3) The rapid approach in the 
vicinity of v/v m = 1 of the differential heats 
for nitrogen on all the carbon surfaces 
studied to values only slightly greater than 
E L , the heat of vaporization of nitrogen, 
provides experimental confirmation of the 
B.E.T. theory since the value of v m is deter- 
mined from the experimental data by a 
method which is entirely dependent upon 
this theory. 

Calorimetric studies with the elementary 
gases oxygen and argon 'produced results 
entirely analogous to those for nitrogen, the 
adsorption of both these gases being ex- 
clusively of the physical type. 

Although the experiments with nitrogen 
already described have served to bring out 
differences among the various carbon blacks 
used, it is obvious that heats of adsorption 
using hydrocarbons as adsorbates would 
provide data which would be more closely 
related to the problem of rubber reinforce- 
ment, in which there must be forces in op- 
eration between the carbon black surface 
and adjacent portions of the large hydro- 
carbon molecule, rubber. For this reason, 
the next phase of the work to be described 
involves calorimetric measurements with 
butane and the butenes as adsorbates at the 
convenient temperature 0°C. Because each 
of these substances boils within a few de- 
grees of 0°C, the p value for each is in the 
vicinity of one atmosphere pressure, and it 
is convenient to work over the relative pres- 
sure range in which the monolayer is com- 
pletely filled. The results of these experi- 
ments are summarized in Fig. 4. In this 
figure are given the differential heat curves 
for butane, 1-butene, cis-2-butene, and 
trans-2-butene on Grade 6 carbon black. 
It is to be noted that the general shape of 
the heat curves is similar to that of nitrogen 
at —195° with the differential heats ap- 
proaching E L after the completion of the 
statistical monolayer. It is noteworthy that 
there is no great difference in the heats of 
adsorption of the butane and the unsat- 

urated hydrocarbons. This evidence ap- 
pears to indicate that the unsaturated sec- 
tions of the rubber molecule are not neces- 
sarily the points at which the rubber is 
bonded to the carbon surface, and that the 
adsorption process and possibly also rubber 
reinforcement involve forces which are pri- 
marily physical in nature. 

As this talk has progressed, I think it 
must have become abundantly evident that 
those of us who have done the heats of ad- 
sorption work at Amherst are very greatly 
indebted to Dr. Brunauer and his co- 
workers for the pioneer work they have done 
in the field of adsorption, work which has 
made our experimental results considerably 
easier to interpret as a contribution to the 
fundamental problems of adsorption as well 
as to the application to rubber reinforce- 


(1) Brunauer. The adsorption of gases and 

Vapors, Vol. 1: Physical adsorption. 
Princeton, 1943. 

(2) Bruxauer, Emmett, and Teller. Journ. 

Amer. Chem. Soc. 60: 309. 1938. 

(3) Emmett and Brunauer. Journ. Amer. 

Chem. Soc. 59: 310. 1937; 59: 1553. 
1937; 62: 1732. 1940. 

(4) Beebe and Stevens. Journ. Amer. 

Chem. Soc. 62: 2134. 1940. 

(5) Beebe, Biscoe, Smith, and Wendell. 

Journ. Amer. Chem. Soc. 69: 95. 
1947; Beebe, Polley, Smith, and Wen- 
dell. Journ. Amer. Chem. Soc, forth- 

(6) This high temperature treatment would of 

course have the effect of removing any 
volatile matter from the surface of the 
Grade 6 carbon as well as changing the 
degree of crystallinity of the underlying 
solid. Our calorimetric experiments on a 
sample of Grade 6 black "devolatilized" at 
927°C make it more plausible to conclude 
that the marked difference in the energy 
states of surface between the Grade 6 
black and the Graphon is due to an al- 
teration in the underlying solid by the 
high temperature treatment rather than 
to an}^ chemical change in the super- 
ficial la vers of the pigment (see reference 

(7) Details of the experimental method, as 

well as data on several other blacks and 
a more extensive discussion are to be 
found in the publications cited in refer- 
ence 5. 

Apr. 15, 1948 



METEOROLOGY. — Loose usage of weather words. 1 W. J. Humphreys. 

Fog, haze, visibility, cyclone, blizzard, 
sleet, drought — what are they? There is no 
excuse for so using words that they convey 
no idea whatever or, at most, only vague 
suggestions of an idea. But frequently they 
are so used, and for several reasons. The 
usual reason is the ignorance of the speaker 
or writer, and the next most common, per- 
haps, the fact that few indeed practice that 
best of all courses in rhetoric: "Have some- 
thing to say; say it; quit talking about it." 
Of course, an unwilling witness may be 
consciously vague, and sometimes justly so, 
but most of us do a lot of talking in our 
sleep, as it were. And the subject we discuss 
in vaguest terms is that which concerns us 
most, and about which we have talked the 
most since that time when certain things 
happened in the Garden of Eden "in the 
cool of the day" — the weather. Even the 
term "weather" itself is used in the vaguest 
sort of way, as anyone will realize as soon 
as he tries to define it accurately. 

Not only are we nearly always vague 
when speaking about weather as a whole, 
but also, as a rule, we are equally inexact 
when speaking of any of its elements, or 
of other phenomena dependent upon or 
associated with it. For example, fog and 
haze are mixed up in a scandalous way 
despite the great and growing importance 
of each to the aviator. They often are used 
as synonymous terms, and again as though 
a fog were just a dense haze and a haze 
merely a light fog. Now, they are not the 
same thing at all and need not often be con- 
fused with one another. A true fog is a cloud 
of water droplets in the space immediately 
above the surface of the earth — a cloud on 
the earth. A haze, on the other hand, is a 
cloud of dust particles of whatever origin, 
such as impalpable sand caught up by 
desert winds, the smoke of forest fires, 
pollen from pine clad mountains, and the 
like. It may, and commonly does, extend 
down to the surface of the earth, but it does 
not always do so. Sometimes, too, its upper 
surface is as sharply limited and- as clearly 
visible in the direction of the horizon (not 
vertically) as is that of a waveless ocean. 

1 Received October 10, 1947. 

Each decreases visibility. That is why the 
aviator is so concerned with them. Perhaps, 
too, as they bother him in much the same 
way he may feel justified in calling either or 
both of them by whichever bad name hap- 
pens to come to his mind first. Indeed they 
often are so confused, but they do not need 
to be, for it is quite easy to distinguish 
between them. The fog droplets are so large 
that they reflect equally, or nearly so, 
lights of all colors. It therefore appears 
white. A large portion of the haze particles, 
on the other hand, are so small that they 
scarcely reflect light at all, but just diffuse 
or scatter it. Furthermore, they scatter the 
blue, or short-wavelength light, to a far 
greater extent than they do the longer 
wavelength, or red light. This being so, the 
glare that results from the scattered light, 
and which so greatly reduces visibility, can 
largely be prevented from reaching the eye 
by the use of amber, or red, glasses. Clearly, 
then, the cloudlike obstruction is a dust 
haze whenever it has a bluish color, and 
whenever visibility through it is improved 
by the use of a red or amber screen. Simi- 
larly, it almost certainly is a fog when it is 
white and equally opaque to all colors. 

Not only is it generally easy to make this 
distinction, but frequently it is important 
to do so, because a fog is likely soon to dis- 
appear by evaporation, while a haze hangs 
on until washed out by rain, thinned by 
convection, or blown away by clear air. 

Visibility. — We have just been glibly 
using a rather new weather term, visibility, 
that needs some explanation. It seems to 
have an obvious enough meaning until we 
try so to define it that one visibility can be 
numerically compared with another. Then 
the trouble begins, for it is hard to say how 
many times one object is more visible than 
another. In fact for practical purposes Ave 
never define visibility that way, but just 
say that it is the distance to which objects 
can be seen, or, better, the greatest distance 
to which objects of appropriate size can be 
recognized by a person of average but un- 
aided vision. Evidently visibility is not a 
distance at all, though dependent on it. 
Nevertheless, this term with its forced 
definition is very useful and seems destined 



to hold indefinitely an honorable place 
among the innumerable and growing host 
of weather words. 

It is worth while also to note that there 
are two quite differently caused visibilities, 
night visibility and day visibility. The 
former may be defined as the greatest dis- 
tance at which a standard light can be seen 
by a normal, unaided eye. This distance ob- 
viously is fixed by the rate of depletion of 
the light on its way to the observer, since 
perception fails in this case owing to the 
feebleness of the light received. Day visibil- 
ity, on the other hand, is limited mainly by 
an entirely different cause — not chiefly by 
depletion of light but by addition of light. 
When the fog or haze between the observer 
and distant objects reflects or scatters so 
much light from the sun, or other sources, 
as to appear luminous the contrasts of 
light, by which outlines are perceived and 
things recognized, soon are lost in the gen- 
eral glare. To repeat, for it is an important 
distinction, the brighter the light the farther 
it can be seen. The brighter the day the 
w r orse the visibility — the greater the blind- 
ing glare. 

Fog and cloud. — Not only is fog often 
confused with haze, to our occasional an- 
noyance, but frequently, too, it is con- 
founded with cloud, and the second of these 
confusions is the more difficult of the two to 
disentangle. We may say, and it is quite 
true, that a fog is a surface cloud formed by 
surface conditions. But what then shall we 
call what is left of it when the under portion 
is burned off, as so commonly happens to 
sea fog when it drifts in over land? Shall we 
call it high fog, lifted fog, veillo cloud (the 
name given to it in southern California), 
or stratus cloud? All these terms are in good 
and constant usage, but "stratus cloud" is 
the best, for that is what the erstwhile fog 
has now become. Wherever, though, it con- 
tacts with the surface, whether water, hill 
or mountain, it should still be, and generally 
is, called fog. That is, the same sheet of 
water droplets is called cloud at one place, 
where it does not extend to the surface, and 
fog at another, w T here it does. 

Name of cloud. — And then w r hen we are 
sure the thing we observe and want to talk 
about, or record, really is a cloud and not a 

fog, what sort of a cloud shall we call it, 
for there are several kinds to which disl inct 
names are given? Way back in the year 
1803 an English chemist, or apothecary, as 
we would say in America, Luke Howard, 
gave Latin names to several distinct kinds 
of clouds, names so apt that they soon 
came into general use in all countries. Later 
on a few other names were added to the 
original list until all the more common varie- 
ties of clouds had each a distinct designa- 
tion that seemed entirely appropriate and 
was clearly defined. But new cloud pictures 
soon were needed, the old charts being out 
of print, and this offered an opportunity to 
revise names and definitions as well. So a 
committee was appointed for that purpose, 
a committee that at once began "throwing 
monkey-wrenches," or so some think, into 
the erstwhile smoothly running cloud ma- 
chine. We used to be quite sure, for in- 
stance, that a nimbus cloud was one from 
which rain or snow was falling in quantity, 
or looked as though precipitation from it 
was imminent. This was pretty satisfactory, 
and many of us want to keep the good we 
have. But no, we are told by this committee 
that those who feel that they really must 
retain this life-long name may be graciously 
permitted to do so provided they pledge 
to mean thereby only that insignificant 
little cloud that drifts along in the rain 
somewhere between the surface of the earth 
and the base of the heavy cloud out of winch 
the rain is falling. That is, this ancient and 
honorable name must go out of use alto- 
gether, or else be given to that little, unim- 
portant ragamuffin the sailor calls "scud." 
This is not clarification. It is needless con- 

Then, too, w T e felt as certain as one could 
about anything that the good and appropri- 
ate name "cumulo-nimbus" had come to 
stay and to mean a cumulus (wool-pack or 
heaped-up) cloud from which rain was fall- 
ing, and in which thunder and lightning 
nearly always occur. No. We now are asked 
to restrict this name to the cumulus cloud 
whose top has been, or is in the process of 
being, drawn out into a thin, fibrous sheet. 
And this in spite of the fact that often a 
thunderstorm cloud, a cumulo-nimbus as 
heretofore known, may give much rain and 

Apr. 15, 1948 



yet produce no fibrous sheet at all; and in 
face, of the further fact that occasionally a 
cumulus cloud may produce no rain, but, 
for all that, a high sheet of fibrous cloud. 
For generations we have been accustomed 
to consider the falling of rain from the base 
of a cumulus cloud the necessary and suf- 
ficient condition for calling it a cumulo- 
nimbus. Now we are asked to note whether 
or not a fibrous cloud is developing out of its 
top — certainly a radical change of meaning, 
and a wholly unnecessary one. 

It would be well, however, for the aviator 
to remember that, in middle latitudes, a 
cumulus cloud that is developing a high 
fibrous sheet very probably, though not cer- 
tainly, is also, and at the same time, giving 
rain below with thunder and lightning ; and 
that the region under a cumulus that has 
not yet begun to display such a sheet, pre- 
sumably, but not surely, still is free from 
rain or other disturbance. To him it is a 
very real warning flag. Nevertheless, this 
fact does not in the least justify this change 
of name, a change that is only confusion 
confounded — or the other way around. 

The next monkey-wrench disarranged the 
alto-stratus cloud. This term used to mean 
just what it says — a high layer cloud; that 
and nothing more. Now we are asked to use 
this term only when portions of the cloud 
show some fibrous structure. Here, too, is 
only confusion, for those who insist on this 
definition insist, too, that this is the rain- 
producing cloud. That is, we must call that 
wide-spread, lightning-free cloud, from 
which rain is falling abundantly, alto-stra- 
tus, and at the same time we must not call 
it alto-stratus unless we can see in it one or 
more fibrous patches — thread-bare por- 
tions, consisting, we believe, of snow parti- 
cles. Truly, this puts us in much the same 
puzzled frame of mind as was Pat when the 
doctor told him to take one pill three times 
a day! 

The cirro-cumulus cloud is another vari- 
ety that many of us will insist that we 
know when we see it, since it is just a field 
of many little cloudlet balls and ripples (a 
"mackerel sky" if in orderly rows, a "cur- 
dle sky" if the cloudlets are numerous and 
without order) very high and too tenuous to 
show evidence of shading, or hide the sun. 

No, again says our authority, this thing 
that is a cirro-cumulus cloud must not be 
called cirro-cumulus unless it has been seen 
to be formed from a cirrus cloud or cirro- 
stratus, that is, from a thin or relatively 
dense cloud, respectively, of fibrous form. 
Said the yeoman to a yokel: "That is a 
fine hog of mine over there in the barn 
lot." "I am not so sure it is a hog," said the 
yokel. "Why not?" "Oh, I didn't see him 
grow up from a pig." "Damned fool." 

Humidity, a term that comes to mind 
when we are talking about clouds, is one of 
the most vaguely used of all weather words. 
Most of us realize that water is somehow 
or other involved in its meaning, but just 
how, when we speak of the air as being 
humid, is not always clear. Certainly the 
air is not wet in the sense that our hands 
become wet when we wash them. Perhaps 
we can run the trouble doAvn this way: 
Water can and does exist in the gaseous 
state as well as in the liquid and solid 
states. Furthermore, the amount of water 
that can occupy a given space in the form 
of a gas, for instance the amount that is in 
a cubic inch of the "empty" space in a 
tightly closed bottle containing some water, 
rapidly increases with increase of tempera- 
ture. Strange as it may seem, too, this 
amount is not appreciably affected by the 
presence of other gases, whether added 
singly or in whatever combination. 

From these facts it is evident that the 
expression "humidity of the air," or "hu- 
midity," for short, can mean any one of at 
least three different things. To be under- 
stood, therefore, one must specify which 
particular humidity he is talking about. If 
he just says "humidity" he will be nearly 
as badly off as the fellow who called for 
Jones at a Welsh college. "Ah Jones" — 
most of the windows went up. "Tom 
Jones" — half of the windows stayed up. 
"I mean the Tom Jones that has a tooth- 

If one means, as often is meant, the 
amount or weight of the water vapor in the 
air per unit volume he must say "absolute 
humidity." If, however, as is more often 
the case, one means the ratio of the amount 
of water vapor actually present per unit 
volume to the greatest amount that could 



exist in the same volume at the same tem- 
perature — the ratio, in brief, of the actual 
quantity to the saturation quantity — it is 
necessary to say "relative humidity." 
Finally, we sometimes mean the weight of 
the water vapor per unit weight of the moist 
air, in which case the proper expression is 
"specific humidity." The term "humidity," 
as qualified in any one of the above three 
ways, has a definite and useful meaning; 
but as commonly used in weather talk it 
seldom conveys a clear idea to the hearer, 
and rarely contains a definite concept when 
it leaves the speaker. Generally, such idea 
as is associated with this term starts mud- 
dled and ends fuddled. 

Wind direction. — Here is another source 
of confusion, one from which force of mem- 
ory alone can protect us, the naming and 
charting of wind directions. The weather 
man's wind vane generally is in the form of 
an arrow of some sort, and so constructed 
that the point of the arrow always turns to 
face the wind. In other words it flies against 
the wind. On his maps and charts, however, 
he gets all his little wind arrows turned the 
other way around, for here they fly not 
against the wind but with the wind. What, 
then, does he mean when he says "north 
wind," for instance? That is a toss-up until 
he tells us that he always means a wind from 
the north. Similarly, by "east Avind" he 
means a wind from the east, by "south 
w T ind" a wind from the south, and so on for 
every point of the compass. It is all simple 
enough — if you don't forget. 

Veering and backing. — Not only the word 
used to designate the direction of a wind 
can be confusing, as just explained, but 
also, and to an even greater extent, those 
commonly employed to specify the order of 
its change of direction. Whenever the wind 
at a particular place so shifts, or changes in 
direction, as to cause the wind vane to 
turn clockwise we say that it is veering, 
and when it so changes as to cause the vane 
to turn counter clockwise we say it is 
backing. This, too, is very plain and easy, 
if only we could remember it — if we could 
keep from getting like the old lady who said 
she knew that good eggs sank or swam, 
but had forgotten which. Indeed it is even 
worse than that, for while, as used by most 

writers, backing is backing, and veering is 
veering, the world over, others reverse the 

terms with change of hemisphere, calling 
that shifting of direction veering, in the 
southern hemisphere, which, in the northern 
they call backing; and in the southern 
backing, that which, in the northern, they 
call veering. Surely the way some words 
are used is perplexing. 

Surface wind is another of the vague 
terms often used in discussions of the move- 
ments of the air. The trouble comes from 
the uncertainty as to where and what the 
"surface" is. In the case of calm water there 
is not much doubt about where the surface 
is, at least in the every day practical use of 
the term surface. Over land, however, we 
usually find a greater and greater entangle- 
ment of grass and other vegetation before 
reaching the soil, and therefore are quite 
unable to say just where the "surface" is. 
But in any case where there is a definite sur- 
face, as over a lake, for example, the air 
exactly at the surface we believe to be calm, 
whatever the wind at appreciable heights 
above that surface. It would seem then that 
by the expression "surface wind" we 
should always mean no wind — no horizon- 
tal movement of the air. However, this 
strictly face value of the expression is not 
at all what we do mean when we use it. We 
mean the wind, not at the surface, as the 
expression implies, but at some appreciable 
height above it, namely at the level at 
which the velocity of the air is measured — 
a level that in practice may be anything 
from a few feet to tAvice as many hundred 
feet above the ground. Clearly then, since 
the velocity of the Avind increases rapidly 
Avith increase of height, and since hills and 
hollows, trees, buildings, and every other ir- 
regularity of, and on, the ground affect the 
velocity of the loAver air, the term "surface 
Avind," like the expression "quarter of a 
lot," has the semblance of meaning some- 
thing definite but only the semblance. 

Fair. — Certainly everyone should knoAV 
the exact meaning of this term, but many 
do not and refuse to learn. We may use as 
A^ague terms as Ave like AAiien merely making 
talk about the Aveather, but the few terse 
sentences used by the forecaster certainly 
should be clearly and correctly understood 

Apr. 15, 1948 



by all who read or hear them. Indeed they 
are clearly understood for they are very 
carefully constructed to that end, but un- 
fortunately they are by no means always 
correctly understood. The chief confusion 
arises from a single one of the forecaster's 
terms, "fair" — his favorite word (formerly, 
at least) if one might judge from the fre- 
quency of its occurrence. The trouble here 
comes from the fact that the forecaster and 
a large portion of his audience, that is, the 
general public, attach entirely different 
meanings to this familiar word. He means 
fair weather; they mean fair skies. To him 
it is the antonym of foul, and means 
weather suitable for outdoor occupation; 
to them it is the antonym of clouds and im- 
plies abundant sunshine. 

What to do about it is the question. It is 
precisely the word to use, or certainly 
would be if rightly understood, but unfor- 
tunately there are many who do not know 
what it means in this connection, and who 
even are unwilling to learn — who resent 
being told, who insist that they know what 
words mean without looking them up in a 
dictionary or having some smart Aleck tell 
them. It is too bad, this confusion of mean- 
ings, and the worst of it is, there is no ob- 
vious and simple way to make the matter 
any better. 

Cyclone. — This is another weather term 
that often fails to carry the meaning in- 
tended. Many people call the "twister" a 
cyclone, that smallest but most violent, 
freakish and destructive of all storms. The 
meteorologist calls this madly whirling devil 
a tornado, but the man from Missouri, or a 
neighboring State, comes back with em- 
phasis: "What are you talking about? If 
my cyclone cellar isn't a place for dodging 
cyclones, then you will have to show me." 
This much of the confusion comes mainly 
from giving the same name, "cyclone," to 
two entirely different kinds of storms, and 
two different names, "cyclone" and "tor- 
nado," to the same sort of disturbance — 
the same name to two things, and two 
names to the same thing. 

But this is not all the confusion. If you 
turn to the word "cyclone" in some large 
dictionary, or even a meteorological vo- 
cabulary prepared by experts, you are likely 

to find its definition to be: "An area of low 
atmospheric pressure," or some similar ex- 
pression. Now, this is inexcusable confusion. 
A cyclone is not an area at all, any more 
than a house is an area. It is first and fore- 
most a system of winds, and secondly, 
characterized by such and such particulars 
(extensive and about a center of low pres- 
sure, accompanied by clouds and wide- 
spread precipitation, et cetera) as suffice to 
distinguish it from all other systems of 
Avinds; just as "man" is first of all an ani- 
mal, and, secondly, possessed of certain 
qualities, such as rationality, risibility, or 
whatnot, that belong to no other animal. 
This is just a fine example of the innumer- 
able cases in which we neither say what we 
mean nor mean what we say. 

Anticyclone. — By its very name one 
would expect the anticyclone to be some- 
thing quite the reverse of the cyclone. And 
so it is in several respects. It occupies a 
region of relatively high atmospheric pres- 
sure, not low, as does the cyclone; its winds, 
like those of the cyclone, are directed 
spirally about its center, but outward, not 
inward, and in the opposite sense; it is at- 
tended, usually, by clear skies, not over- 
cast, and fair weather not foul. Meteoro- 
logically the cyclone and the anticyclone 
are distinctly antithetical. Lexicologically, 
however, they usually have one important 
feature in common, for generally each is er- 
roneously defined as an area, and not cor- 
rectly defined as a particular system of 

Secondary. — If one were in a teasing 
frame of mind and wanted some fun with 
a meteorologist, he hardly could find a bet- 
ter way to succeed than by asking him what 
a secondary is. He uses this term a lot, and 
by it he always means a cyclonic storm. 
Likely as not he will tell you that a second- 
ary cyclone is any one that is not a pri- 
mary; and that a primary cyclone is any 
one that is not a secondary. Often, and es- 
pecially in certain regions, there develops on 
or near the outer border of a system of C3 r - 
clonic winds, and doubtless incident to 
them, a similar system, at first of relatively 
small extent and strength, but which later 
grows in size and intensity until, in many 
cases it itself becomes the main storm, and 

J 28 


occasionally even the only one. This storm, 
at least in its earlier stages, is often called a 
secondary cyclone, or secondary, for short, 
But other storms of less certain history also 
have been called secondaries for reasons 
that sometimes seem to be known only to 
the perpetrator — and he never tells. Yes, 
ask a meteorologist about secondaries and 
pretty soon you will have him in a corner, 
if not up a tree. 

Blizzard, a good example of things con- 
fused, is a fine word, even if its pedigree is 
unknown, so long as it means a cold, driving 
wind filled with blinding snow — a sure-to- 
goodness storm of the Plains. But how un- 
worthy of itself, how fallen, how decrepit 
and addled it is when made to mean only a 
little snow flurry that even a baby scarce 
would notice. A real blizzard commands re- 
spect and deserves a strong name all its 
own, but the little parlor things that in the 
Eastern States we so often braggingly call 
blizzards are not worthy, in comparison, 
even to be called "blizzetts." A blizzard 
used to be a blizzard with no doubt about it; 
now it is anything that has a little snow in 
it, but one never knows exactly what. 

Hail. — Here is a mix-up, too. As used by 
the U. S. Weather Bureau, and by many 
people all over the country, only lumps of 
ice that fall in thunder storms are called 
hail. Lots of others, though, will not hear of 
this restriction, but insist that the little ice 
pellets that fall only in the winter shall also 
be called hail. That is what they call them, 
as of course they have a perfect right to do, 
despite the confusion it causes, and there is 
no help for it, swear as you may. 

Sleet — Hail is bad enough in its confu- 
sion, but sleet is worse, for it has three en- 
tirely distinct and much-used meanings. As 
sung by the poets, and as recorded by the 
U. S. Weather Bureau, sleet is that which 
rattles against the windowpane and nothing 
else. It is frozen raindrops and occurs only 
in winter and when the temperature of the 
lower air is below the freezing point while 
that of the air at around 500 to 1,000 feet 
elevation is distinctly above it. The rain 
that falls from the warmer air is frozen by 
the colder layer below, and reaches the sur- 
face in the form of small roundish pellets — 
ice shot. This is sleet, according to one defi- 

nition and extensive usage. According to 
British usage, though, and the custom of 
many in America also, sleet is a mixture of 
rain and snow. They call the frozen rain- 
drops hail, or winter hail. This is confusion 
enough, but by no means the whole of it, 
for the engineer, ignoring both the above 
definitions, restricts the name sleet to the 
smoothish coats of ice that sometimes form 
on wires, street-car rails and other exposed 
objects. This sleet, in the engineer's sense 
of the term, the U. S. Weather Bureau calls 
glaze. It is the characteristic feature of an 
ice storm. 

And nothing can be done about it. Even 
if we could get the engineer to call glaze 
(an excellent word) that which he now calls 
sleet and restrict this term, as so many of us 
do, to the stuff that rattles, there still would 
remain the mixture of snow and rain to re- 
christen. Of course, a word-telescoping 
genius like Lewis Carroll might call it 
"snane," but then some one would be sure 
to laugh at it, if he did not even make a face. 

Ice-flower. — If sleet is a vague term owing 
to its three definitions, what shall we say 
about "ice-flower" that means any one of 
five distinct things? Two or three genera- 
tions ago Tyndall tried the happy experi- 
ment of putting a sheet of ice in the usual 
position of a lantern slide, and was rejoiced 
to see blossom on the screen a beautiful 
bed of six petaled flowers — images of course, 
of internal melted ice crystals. These 6- 
rayed cavities Tyndall, in his poetic way, 
and very appropriately, called ice-flowers. 
Pretty soon some one else, also with a poetic 
fancy, gave the same name to the beautiful 
fernlike figures Jack Frost traces on your 
windowpane. Then some nature lover, en- 
thusiastically describing the tufts of frost 
that sometimes in bitterly cold weather 
grow up numerously on a sheet of ice, called 
them also "ice-flowers. " Even the little 
columns of ice that spring up from damp 
soil, and resemble in a measure the familiar 
Indian pipe, have been called ice-flowers. 
Finally, at least finally up to the present, 
the exquisite, curling and satiny ice ribbons, 
as thin and broad as the blade of a case 
knife, and often fully as long, that in early 
winter grow out from the dead stems of rock 
mint, likewise have been called ice-flowers. 

Apr. 15, 1948 



And no wonder, for among the brown leaves 
of the woods these ice formations look for 
all the world like a field of beautiful white 

In every one of these five cases the name 
ice-flowers seems appropriate, yet its use 
for more than one, preferably the first, is 
to be deprecated, for it leads only to con- 

Climate. — It is a long call in one sense 
from ice-flowers to climate, but the terms 
have one bad feature, muddle meaning, in 
common. Perhaps climate is most often de- 
fined as "average weather." That would 
seem to justify defining weather as irregular 
climate. However, both these snap defini- 
tions are hopelessly inadequate. The aver- 
age annual temperature of a place, average 
rainfall, and average all-the-other-things do 
not tell the story of its climate. We need 
for this purpose to know also the extremes, 
frequencies of such and such values, normal 
run of the weather elements through the 
year, and a lot of other matters as well. In 
short, the climate of a place is neither its 
average weather nor the average of its 
weather, but the history of its past weather. 

Drought. — This much used term also is a 
term confused. Most of us use it glibly as 
though we knew exactly what it meant 
until, perchance, we need to define it pre- 
cisely whereupon our concept of it begins to 
fade away. We ask ourselves whether it is 
a number of consecutive days without rain 
or snow, and, if so, how many. Then we 
wonder whether the amount of the immedi- 
ately preceding precipitation should be con- 
sidered in our definition, and whether the 
time of the year matters. And if we take a 
certain number of consecutive days without 
rain as our definition of drought for one 
region will that also hold for every other re- 
gion? Would such a drought for New Eng- 
land be also a drought for New Mexico? 
If not, then what is a drought? Most of us 
will agree that one one-hundredth of an 
inch of rain will not break a drought, and if 
it will not then certainly we can not define 
a drought as so many consecutive days 
without any rain at all, though sometimes, 
and for statistical purposes, such a rainless 
spell has been called an absolute drought. 
Perhaps we might better define a drought 

as so many consecutive days without more 
than a specified small amount of rain. But 
this does not avoid the difficulties as to 
season and place. 

Really, drought is a hard term to deal 
with. Possibly the best thing to do when- 
ever accuracy is essential is to give it an 
arbitrary, but reasonably exact, definition 
appropriate to the needs of the occasion. 

Spring. — Who knows when spring begins, 
or any of the other seasons, for that mat- 
ter? If we take Tennyson's dictum that it is 
"when a young man's fancy turns to 
thoughts of love," we must conclude that 
it is a perpetual season! If, on the other 
hand, we take the word at its primitive 
meaning, signifying the season when plants 
spring up, we will be forced to conclude that 
the beginning of spring varies from place to 
place and year to year, and also that it de- 
pends on the kind of plant selected as the 
criterion. We will even conclude that it has 
no beginning in ice covered regions, and 
that it goes on forever in tropical lands. 

This indefiniteness led to the more or less 
general adoption of fixed dates for the be- 
ginning and end of each season. In so doing 
the spice of life was amply preserved for 
surely in the dates of these beginnings there 
is abundant variety. According to popular 
English usage spring begins with the first 
of February. In America we put it a month 
later, first of March. Astronomers, though, 
say it begins with the vernal equinox, which 
may be at any instant through the day, 
usually, but not always, on the twenty-first 
of March, and lasts until the coming sum- 
mer solstice, also a slightly variable date. 
Here are three widely used but distinctly 
different dates for the beginning of spring. 
February 1, March 1, and March 21 or 22, 
spread over a period of seven weeks. And 
so it is with the other seasons, for they follow 
each other at approximately equal intervals 
of time. Clearly, then, "first day of spring," 
"last day of summer," and all others of 
their kind are confusing expressions for to 
equally well informed people they convey 
the concept of distinctly different dates. 

End of twilight. — When the end of twi- 
light comes is another point about which the 
astronomer and the general public have oc- 
casion to differ, though confusion in this 



case is not nearly so bad as it is in regard to 
the beginning of spring. The astronomer 
who, for most of his work, prefers the clear- 
est skies and the darkest hours, says twi- 
light ends only with the last trace of scat- 
tered sunlight in the western heavens. On 
clear nights, the only kind in which he is 
interested, this last glimmer disappears 
when the center of the sun is about 18 de- 
grees below the horizon. This does not suit 
the average person who considers twilight 
to end as soon as it gets too dark for people 
to go about their ordinary outdoor occupa- 
tions. On clear evenings this occurs when 
the center of the sun is about 6° below the 
horizon, or in about one-third the time from 
sundown to the end of astronomical twi- 
light. On cloudy evenings twilight, in this 
work-a-day sense, ends much sooner, but 
there is no fixed time for it — the term is 
vague and often confusing. 

Light and dark of the moon. — Whoever 
follows the foolish occupation of moon- 
farming, of planting things that fruit above 
the ground in the light of the moon, and 
tubers that grow under the soil in the dark 
of the moon, is confronted with the puzzling 
necessity of knowing just when it is light of 
the moon and when dark. And the more he 
tries to be certain of the matter the more 
confused he is likely to become. Some will 
tell him that dark of the moon is that brief 
time, three or four days, before new moon 
when it is not seen at all owing to its near- 

ness to the sun, and light of the moon the 
three or four days centered about full 
moon. Others, while agreeing with this 
definition of dark of the moon, will insist 
that all the rest of the time is light of the 
moon. Still others, with equal assurance, 
will insist that the time the moon is waxing, 
that is, the time of the first and second 
quarters, is light of the moon, and that the 
time of its waning, the duration of the third 
and fourth quarters, is dark of the moon. 
Finally, there are many who recognize light 
of the moon to be all the days when the 
moon is above the horizon most of the fore 
part of the night j and all the rest of the 
time dark of the moon. 

Here are three distinct and widely recog- 
nized definitions of dark of the moon, and 
four of the light of the moon. What then 
can the poor moon farmer do when up 
against such conflicting definitions as these? 
Xobody knows, though it is quite certain 
what he should do — forget the moon fool- 
ishness and plant when the ground is ready 
and the season right, as all sensible farmers 

And these listed above are not the only 
weather words loosely used by the general 
public, whose omnibus excuse is the fact 
that, if restricted to correct understanding 
and clear expression, the pleasant glibness 
of its tongue would be lost in many a dreary 
silence — an honest enough excuse but a 
mighty poor one. 

PALEONTOLOGY. — An interesting occurrence of fossil tracks in West Virginia. 1 
David H. Dunkle, JJ. S. National Museum. (Communicated by C. L. 

Recently, through the generosity of 
Harold T.~Stowell, of Westmoreland Hills, 
Aid., the U. S. National Museum received 
a small slab of rock exhibiting two distinctly 
impressed tracks. The following brief report 
on the specimen is prompted by the proba- 
ble tetrapod nature of the prints and by the 
geologic details of its occurrence. 

The prints are impressed upon the upper 
undulating surface of a thin block of dense, 
finely grained sandstone, composed of many 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received December 
5, 1947. 

small cross-bedded laminae of variegated 
red and buff color. As found b}^ Mr. Stowell 
in 1939, the slab laj^ loose at the base of a 
cliff on the East Bank of the Greenbrier 
River, about 1 mile south of the mouth of 
Island Lick Run, in the Watoga State Park, 
Pocahontas County, W. Va. 

The physical expression of fossil track- 
ways is dependent on a number of variable 
conditions. Seldom are the remains of the 
causative agent found associated with the 
prints. The structure of the contributing 
organ, therefore, must be inferred. This 
latter is very often difficult because the 

Apr. 15, 1948 dunkle: an occurrence of fossil tracks in west Virginia 


completeness of the impression is subject, 
at the time the tracks were made, to the 
composition and the consistency of the sub- 
stratum and, subsequently, to all the mul- 
tiple, special factors for their preservation. 
The interpretation of the present examples 
is not exempted from these general difficul- 

The two tracks, oriented parallel to each 
other, are each composed of the depressions 
made by three, stout, distally tapering dig- 
its. When the block is placed with the di- 
vergent extremities of the impressions di- 
rected away from the observer (Fig. 1), it 
may be seen that the track on the left is 
more deeply impressed and is situated 
slightly above the one on the right. 

From the similar lengths and practically 
identical angles of divergence of the corre- 
sponding digital impressions in each of the 

two tracks, it is conceivable that the prints 
could have been made by one and the same 
appendage of an animal crossing the photo- 
graph transversely. In this latter event, 
however, the limb would seem necessarily 
to have extended out at a right angle to the 
longitudinal axis of the body with little or 
no anterior flexure. The literature on fossil 
trackways fails to reveal the occurrence of 
such a structural condition in the Late 
Paleozoic. Further in opposition to such an 
interpretation is the fact that the distance 
between the two prints would represent an 
extremely short stride, especially when the 
slab is large enough transversely to exhibit 
both preceding and succeeding impressions. 
During the tetrapod propulsive cycle 
(SchaefTer, 1941; Evans, 1946), the body 
weight, first distributed over the entire sur- 
face of either the hand or foot, progressively 


Fig. 1.— Photograph of fossil tracks (U.S.N.M. no. 17656) obtained in the Watoga State Park, 
Pocahontas County, W. Va. Reproduction approx. X3/5. 



shifts to a final concentration on the medial 
digits. While variable (Colbert and Schaef- 
fer, 1947), this action ideally results in a 
deeper impression of the inner side of the 
organ than any other of its parts. The sides 
of the present fossil footprints, which are 
adjacent to each other, are clearly more 
deeply impressed than their distant sides. 
Thus, while the evidences are conflicting, it 
is here assumed that these tracks were made 
by a limb and its complement from the 
opposite side. Whether that pair of append- 
ages was anterior or posterior can not be 

Markings of the pads of either heel or 
palm are not discernible. Impression of the 
digits alone indicates that the tracks were 
made on a fairly firm substratum. As 
pointed out by Colbert and Schaeffer (1947), 
the lateral digits are structurally the most 
divergent and under the least optimum of 
conditions are the most poorly defined in 
trackways. Thus, while 3-toed impressions 
have been encountered in practically every 
known geologic occurrence of tetrapod 
tracks, it remains uncertain whether com- 
plete impressions are being dealt with in 
the present case. No attempt here is made 
to assign these prints to any of the scientific 
names available because of the above men- 
tioned uncertainties of interpretation and 
because of the questionable advisability of 
such practice. The specimen is no less in- 
teresting for this failure, however, because 
an early and unknown animal of considera- 
ble dimensions is indicated. The better pre- 
served left imprint measures roughly 35 
mm across the proximal base of the digits. 
The distance between parallel lines pro- 
jected through the medial borders of the 
prints approaches 62 mm. 

The block of sandstone bearing the im- 
pressions may be assumed to have been de- 
rived at or very near the site of its discovery. 
No evidences of transportation can be ob- 
served. The edges of the slab remain sharply 
angular. Fragments of soft red shale still 
adhere in the concavities on both its upper 
and lower surfaces. Furthermore, its lithol- 
ogy is identical with that of the bedrock 
exposed in the immediate vicinity as well as 

for many miles upstream. Price (1929) iden- 
tifies this rock as the Pocono formation. As 
summarized by Branson (1910) and Colbert 
and Schaeffer (1947), the oldest previous re- 
ports of definitive tetrapod tracks in the 
United States are restricted to the upper 
Mississippian Chester Series. The literature 
(Butts, 1940; Chadwick, 1935; Willard, 
1936) suggests a transgressive character to 
the Pocono sediments and a consequent 
variation in age from place to place. Not- 
withstanding, the Pocono is well below the 
Chester equivalent, the Mauch Chunk, in 
the West Virginia section. The present oc- 
currence, therefore, seems to be the earliest 
yet known from our country and approaches 
in age the oldest authenticated tetrapod 
tracks from the Horton Series of Nova 
Scotia (Dawson, 1894; Sternberg, 1933). 


Branson, E. B. Amphibian footprints from 
the Mississippian of Virginia. Journ. 
Geol. 18(4) : 356-358, 1 fig. 1910. 

Butts, Charles. Geology of the Appalachian 
Valley in Virginia. Pt. 1 : Geologic text and 
illustrations. Virginia Geol. Surv. Bull. 
52: i-xxxii, 1-568, 10 figs. incl. index and 
geol. sketch maps, 10 tables, 63 pis. 1940. 

Chadwick, G. H. What is "Pocono"? Amer. 
Journ. Sci., ser. 5, 29(170): 133-143, map. 

Colbert, E. H., and Schaeffer, B. Some 
Mississippian footprints from Indiana. 
Amer. Journ. Sci. 245(10): 614-623, 1 fig., 
1 pi. 1947. 

Dawson, W. Synopsis of the air-breathing ani- 
mals of the Paleozoic in Canada, up to 1894. 
Trans. Roy. Soc. Canada 12(4): 71-88. 

Evans, F. G. The anatomy and function of the 
foreleg in salamander locomotion. Anat. 
Rec. 95: 257-281, 6 figs. 1946. 

Price, P. H. Pocahontas County. West Vir- 
ginia Geol. Surv. County Repts. : 531, 21 
figs., 71 pis., 2 maps. 1929. 

Schaeffer, B. The morphological and func- 
tional evolution of the tarsus in amphibians 
and reptiles. Bull. Amer. Mus. Nat. Hist. 
78(6): 395-472, 21 figs. 1941. 

Sternberg, C. M. Carboniferous tracks from 
Nova Scotia. Geol. Soc. Amer. Bull. 44: 
951-964, 1 fig., 3 pis. 1933. 

Willard, B. Continental upper Devonian of 
northeastern Pennsylvania. Geol. Soc. 
Amer. Bull. 47(4): 597-599, 3 figs., 3 pis. 

Apr. 15, 1948 



MYCOLOGY. — The swarm-cells of Myxomycetes. 1 Eugene AY. Elliott, State 
University of Iowa. (Communicated by G. W. Martix.) 

Biflagellate swarm-cells have been ob- 
served as occurring occasionally in Myxo- 
mycetes by many observers, but it has been 
generally held that the swarm-cells in this 
group are normally uniflagellate. DeBary 
(1) in 1884 and Vouk (20) in 1911 reported 
biflagellate forms in exceptional cases only. 
Gilbert (6) found one-fourth of the swarm- 
cells of Stemonitis fusca biflagellate. Yon 
Stosch (19) saw biflagellate cells in many 
other species, but none in the single species 
of Stemonitis that he studied. Gilbert (7) 
reported Dictydiaethalium plumbeurn to be 
uniflagellate, but E. C. Smith (17, 18) twice 
within the following year published photo- 
micrographs showing biflagellate swarm- 
cells in the same species. Howard (10) 
termed biflagellate swarm-cells in Physarum 
polycephalum "common." Sinoto and Yuasa 
(16) studied Ceratiomyxa and four species of 
Myxogastres, finding only one flagellum in 
Ceratiomyxa but occasional bi- and even tri- 
flagellate forms in all the others. Yuasa (21) 
likewise foimd bi- and triflagellate cells in 
Fuligo septica. Jahn (11) in 1928 ignored the 
previous reports of biflagellate swarm-cells; 
and in 1936 (12) he criticized the work of 
Von Stosch, insisting that swarm-cells are 
normally uniflagellate and that all biflagel- 
late swarm-cells are anomalies. Karling 
(18) in his general summary of the literature 
pertinent to the relationships between the 
Plasmodiophorales and the Myxomycetes, 
says that "although the majority are uni- 
flagellate, zoospores with two flagella are 
not uncommon. ..." In 1945 Ellison (5) 
reported biflagellate swarm-cells in propor- 
tions varying from 2 to 26 percent for a 
number of species of Myxomycetes but re- 
tained the assumption that the majority are 

It is not inconceivable that an occasional 
biflagellate swarm-cell could occur as an ab- 
normality in an otherwise uniflagellate 
group. But biflagellation has already been 
reported too frequently in the Myxomycetes 
to represent mere abnormality. And on the 
basis of flagellation as reported in other 
groups it is very unlikely that both uni- 

1 Received December 15, 1917. 

flagellate and biflagellate forms would nor- 
mally exist in the same life stage of the 
Myxomycetes. It seems more probable that 
the second, shorter flagellum is difficult to 
see and is frequently hidden. This study was 
undertaken to determine whether this latter 
assumption is not the case. 


Spores of 21 collections representing 11 
species were germinated for these tests. Fol- 
lowing is a fist of the collections, with the 
State and vear of collection: 

1. Arcyria denudata 

2. Dictydiaethalium plumbeurn 

3. Dictydiaethalium plumbeurn 

4. Enteridium rozeanum 

5. Enteridium rozeanum 

6. Enteridium rozeanum 

7. Fuligo septica 

8. Fuligo septica 

9. Lycogala epidendrum 

10. Lycogala epidendrum 

11. Lycogala epidendrum 

12. Oligonema schweinitzii 

13. Oligonema schweinitzii 

14. Physarum polycephalum 

15. Reticularia ly coper don 

16. Reticularia lycoperdon 

17. Reticularia lycoperdon 

18. Reticularia lycoperdon 

19. Stemonitis flavogenita 

20. Stemonitis splendens 

21. Trichia affinis 

The first cultures were prepared in Sep- 
tember 1946, using Reticularia lycoperdon 
(Coll. no. 15, above), Lycogala epidendrum 
(no. 10), and Trichia affinis (no. 21). Abun- 
dant germination was obtained in the first 
attempts with R. lycoperdon and T. affinis, 
but only one of several cultures of L. 
epidendrum was observed to germinate. 
Repeated attempts to germinate cultures of 
the two oldest collections of Enteridium 
rozeanum (nos. 4 and 5) resulted in the ob- 
servation of occasional swarm-cells in no. -4, 
but only very low percentages of germina- 
tion. Other species gave similar results. 

Cultures were made in Syracuse watch 
glasses using distilled water from which the 
traces of toxic minerals were removed with 
powdered charcoal. It was observed fre- 

































Xew York 




West Vir- 









quently that most of the spores placed in 
the culture dish continued to float on the 
surface of the water, never becoming wet. 
This was especially true of those forms hav- 
ing very small spores, such as Enteridium 
and Lycogala. Various wetting agents were 
tried in an attempt to accelerate the wetting 
and increase the germination of these spores. 

Alcohol was the first wetting agent tested. 
Cayley(4) used 20 percent alcohol for wet- 
ting spores of Didymium sp., securing ap- 
proximately 50 percent germination 
whether the spores were in the alcohol "a 
few minutes" or a full hour. Also she used a 
solution of 0.2 percent mercuric chloride in 
a mixture of equal parts of 95 percent alco- 
hol and water, as a combined wetting agent 
and bactericide. The resulting cultures were 
not bacteria-free and germination was poor. 

In the tests here reported a number of 
different dilutions of alcohol were tried, 95 
percent being the strongest and 20 percent 
the weakest concentration used. Lower con- 
centrations had negligible wetting effect. 
After wetting, spores were washed three 
times with centrifuging and cultured as be- 
fore. No germination was obtained from 
spores wetted with -alcohol, even in Reticu- 
laria ly coper don, in which nearly 100 per- 
cent germination had been secured without 
the use of a wetting agent. 

Trisodium phosphate was tried next. By 
experimentation it was found that spores of 
Enteridium rozeanum, which has the smal- 
lest spores of any species used up to the time 
of these tests, would sink immediately in a 
0.5 percent solution and slowly in a 0.2 per- 
cent solution. In preparing cultures using 
trisodium phosphate as a detergent, wash- 
ing was done as when alcohol was the wet- 
ting agent. Cultures of E. rozeanum (no. 5), 
prepared with the use of trisodium phos- 
phate in either 0.5 percent or 0.2 percent 
solution, germinated nearly 100 percent 
within one hour. Swarm-cells had not been 
seen in cultures of this collection before. 

Trisodium phosphate in 0.5 percent solu- 
tion was used as a detergent in preparation 
of cultures of a number of other collections. 
Abundant germination was produced occa- 
sionally in Lycogala epidendrum (no. 10), 
but no consistent germination was secured 
with any species except E. rozeanum, hence 
toxic effects were suspected. 

In the search for an active detergent 
without toxic effects, two commercial de- 
tergents, "Soilax" and "Dreft," were tried, 
both in 0.5 percent solution. Both acted as 
effective wetting agents, but many spores 
were caught in the foam on the "Draft" so- 
lution and could not be reclaimed by centri- 
fuging. Germination in cultures thus pre- 
pared was similar to, and no more satisfac- 
tory than, that obtained with trisodium 

The fact that bile salts greatly lower the 
surface tension of solutions, as exemplified 
in the Hay test for bile in urine (9), inspired 
the testing of solutions of sodium glyco- 
cholate and sodium taurocholate as wetting 
agents for preparation of cultures. It was 
found that either the glycocholate or the 
taurocholate will wet the spores satisfac- 
torily in 1 percent solution, but lower con- 
centrations act so slowly as to be inade- 
quate. In equal concentrations, these two 
salts, or a mixture of them, are equally effec- 
tive as wetting agents. 

Spores of many species of IVfyxomycetes 
that did not germinate at all when prepared 
directly in water or with other detergents, 
germinated when previously wetted with 
either of the bile salts. And all those collec- 
tions which germinated directly in water 
germinated more quickly and in greater per- 
centage when previously wetted with the 
bile salts. Of course, the spores of two dif- 
ferent collections of the same species fre- 
quently show greatly different percentages 
of germination. Similar differences are ap- 
parent when the bile salts are used, other 
conditions being equal; but the differences 
are substantially reduced. 

Germination of spores directly in 1 per- 
cent solution of the bile salts was tried. 
Protoplasts emerged from the spore cases, 
but developed no further. However, a 
technique was developed in which only one 
washing with water is necessary to free 
wetted spores of the detergent. Approxi- 
mately 1 cc of the wetting agent is placed in 
a centrifuge tube and the spores to be cul-. 
tured are added and stirred until wetted, a 
process which usually takes one-half minute. 
Then the solution is diluted to 5 or 6 cc with 
water and promptly centrifuged. The spores 
are then washed once with distilled water 
by centrifuging and are finally cultured in 

Apr. 15, 1948 



distilled water prepared with charcoal as 
described above. 

It was found that in order to preserve the 
flagella of swarm-cells for observation, the 
killing agent used in preparation of material 
for microscopic examination must act very 
quickly. Smears prepared by air-drying, or 
even by drying as quickly as possible over 
mild heat, as is done in the preparation of 
bacterial mounts, showed recognizable 
swarm-cells, all of which, however, com- 
pletely lacked flagella. 

The most satisfactory results for tem- 
porary mounts were obtained by killing and 
staining on the slide with a drop of iodine- 
potassium iodide solution. For this purpose 
Gram's iodine is used without dilution, den- 
sity of staining being controlled by varying 
the proportions of culture solution and io- 
dine solution which are mixed on the slide. 
Whatever proportions are used, mixing 
must be accomplished quickly and thor- 
oughly if the flagella are to be preserved for 

Permanent slides were prepared for obser- 
vation of flagella by use of a modification of 
the Loeffler stain for bacterial flagella (3). 
The mordant and stain were prepared as 
directed, but the times of application of 
both mordant and stain were reduced from 
five minutes to one-half minute. Smears 
were prepared by various means. Slides were 
thinly coated with albumin fixative, on 
which a drop of culture solution was placed 
and inverted over osmic acid fumes. This 
was then allowed to dry in air. Other 
smears, killed over osmic acid, were fixed by 
heating gently after drying. Still others 
were killed with iodine, as was done in the 
preparation of temporary mounts, and al- 
lowed to stand until the iodine had sub- 
limed. These slides were fixed over heat. The 
crystals of potassium iodide were dissolved 
off in distilled water before staining. 
Equally satisfactory results were obtained 
with all these methods. The iodine method, 
being the simplest, was used. 


Some biflagellate swarm-cells were ob- 
served in every culture prepared of every 
species studied. In some species, notably 
Lycogala epidendrum, Oligonema schweinit- 

zii, and Fuligo septica, the proportion 
of swarm-cells obviously biflagellate was 
nearly 100 percent. In one culture of Dicty- 
diaethalium plumbeum, many zygotes with 
four flagella, in two pairs of two each, were 

In all species except Stemonitis splendens, 
the two flagella were of greatly different 
length, the shorter one being sometimes 
scarcely more than lju long. The length of 
the shorter flagellum is quite constant for a 
given species. The flagella of Stemonitis 
splendens averaged 16^ for the longer flagel- 


Figs. 1-8. — Outlines of representative swarm- 
cells: 1, Dictydiaethalium plumbeum (a-b, swarm- 
cells; c, zygote) (all others are swarm-cells); 2, 
Enteridium rozeanum; 3, a-b, Fuligo septica; 4, 
Lycogala epidendrum; 5, Oligonema schweinitzii ; 
6, Physarum polycephalum ; 7, Reticularia ly- 
coperdon; 8, Stemonitis splendens. 



lum and 14/x for the shorter. Ellison drew 
swarm-cells of Stc?nonitis fusca and 
S. fcrruginea with two flagella approxi- 
mately equal in length, though the other 
biflagellate forms he records have one flagel- 
lum much shorter than the other. Similarly, 
Gilbert (6) shows S. fusca with two nearly 
equal flagella, but all other forms definitely 

The second flagellum, in addition to being 
very short, is usually recurved so as to be 
almost indistinguishable from the outline of 
the cell itself. Even in Siemonitis splendens, 
in which both flagella are relatively long, 
one is usually trailing, so that it is easily 
overlooked. The significance of this point is 
borne out in the following observation: A 
temporary mount was prepared from a cul- 
ture of swarm cells of Dictydiaethalium 
plumbeum. The mount was killed and stained 
with Gram's iodine. Using the 90X oil im- 
mersion objective, the microscope was fo- 
cused upon a swarm-cell in which both 
flagella were easily seen. Then, by touching 
the edge of the cover slip with a dissecting 
needle, the mount was disturbed while the 
original swarm- cell was kept in view. This 
swarm-cell was observed to roll over and 
again come to rest. In its new position, the 
shorter flagellum could not be seen with any 
manipulation of the microscope. A second 
swarm-cell which presented only the longer 
flagellum to view was found, and by simu- 
larly disturbing the cover slip the swarm- 
cell was maneuvered until the shorter flagel- 
lum could also be seen. This same demon- 
stration of the fact that the shorter flagel- 
lum may be — and frequently is — hidden by 
the body of the cell, was also performed on 
mounts of Fuligo septica and Arcyria de- 

The second flagellum is clearly visible on 
only a very few swarm-cells in most mounts. 
Some mounts stained with Loeffler's tech- 
nique were destained to transparency with 
acid alcohol. On swarm-cells thus prepared 
the flagella remained clearly stained, but the 
body of the cell was sufficiently destained so 
that the nucleus and other cellular details 
were visible. When swarm-cells are properly 
stained by this method, careful focusing on 
the anterior portion of the body of the cell 
will reveal a dark line reaching backward 

from the apex to the base of the conical 
anterior portion. Occasionally this dark line 
will be found lying across the conical por- 
tion, either on or underneath the body of 
the cell, but most frequently it is barely dis- 
tinguishable from the outline of the cell. It is 
not found on those swarm-cells on which the 
second flagellum is clearly visible apart from 
the body of the cell. It is believed that this 
is the shorter flagellum which ordinarily is 
closely appressed to the anterior portion of 
the cell. 


Spores of 21 collections representing 11 
species of Myxomycetes were germinated. 
The flagellation of the swarm-cells of these 
species was observed in temporary and 
permanent preparations. In three species, 
Fuligo septica, Lycogala epidendrum, and 
Oligonema schweinitzii, it was possible to 
see two flagella on nearly every swarm-cell 
observed. In all other species, two flagella 
were clearly visible on some swarm-cells, 
and it was demonstrated by manipulation 
of fluid mounts that the second flagellum 
could be brought into view even though not 
originally visible. From this it is inferred 
that if adequate technique were used, the 
second flagellum would be found on all 
myxomycete swarm-cells. 

All of the biflagellate swarm-cells ob- 
served in this study were heterocont. In all 
species except Stemonitis splendens the two 
flagella differ greatly in length. In S. 
splendens the difference is slight, but is 
nevertheless constant. 

The Myxomycetes and the Plasmodio- 
phorales have long been considered related 
groups by may investigators. Formerly, 
the reported existence of anteriorly uni- 
flagellate zoospores in the reproductive cy- 
cle of both groups was regarded as strong 
evidence of this relationship. However, Led- 
ingham (14, 15) showed that the zoospores 
of the Plasmodiophorales are anteriorly bi- 
flagellate, the second flagellum being very 
short and, hence, easily obscured. As a re- 
sult of Ledingham's finding, the supposed 
difference in flagellation was thought by 
many to emphasize a separation between the 
two groups. 

The existence of two blepharoplasts has 
been reported in several species of Myxomy- 

Apr. 15, 1948 friedmanx: a collection of birds from Eritrea 


cetes {5, 8, 19). Bessey (2) regards the 
second blepharoplast as a vestige of the 
biflagellate condition, indicating that the 
Myxomycetes and the Plasmodiophorales 
have arisen from a common ancestor. He 
considers the loss of the second flagellum as 
evidence that the Myxomycetes are of 
higher phylogenetic position. 

Insofar as flagellation is of phylogenetic 
significance, the existence of the second 
flagellum in the swarm-cells of Myxomy- 
cetes as demonstrated by this study, may 
indicate a closer relationship with the Plas- 
modiophorales than has recently been sup- 


(1) Bary, A. de. Yergleichende Morphologie 

und Biologic der Pilze, Mycetozoen, und 
Bacterien. Leipzig, 1884. 

(2) Bessey, E. A. Some problems in fungus 

phytogeny. Mycologia 34: 355-397. 

(3) Buchanan, E. D., and Buchanan, R. E. 

Bacteriology, ed. 2. New York, 1931. 
(Pp. 153-154.) 

(4) Cayley, Dorothy M. Some observa- 

tions on Mycetozoa of the genus Didvm- 
ium. Trans. Brit. Myc. Soc. 14: 227- 
248. 1929. 

(5) Ellison, Bernard R. Flagellar studies 

on zoospores of some members of the 
Mycetozoa, Plasmodiophorales, and Chy- 
tridales. Mvcologia 37: 444-454. 

(6) Gilbert, F. A. On the occurrence of bi- 

flagellate swarm cells in certain Myxo- 
mycetes. Myeologia 19: 277-283. 

(7) . Feeding habits of the swarm cells 

of the Myxomycete, Dictydiaethalium 
plumbeum. Amer. Journ. Bot. 15: 
123-132. 1928. 

(8) Gilbert, Henry C. Critical events in the 

life history of Ceratiomyxa. Amer. 
Journ. Bot. 22 : 52-74. 1935. 
(9) Hawk, Philip B., and Bergeim, Olaf. 
Practical physiological chemistry, ed. 11. 
Philadelphia, 1937. (P. 653.) 

(10) Howard, Frank L. The life history of 

Phvsarum polvcephalum. Amer. 
Journ. Bot. 18: 116-133. 1931. 

(11) Jahn, E. Myxomycetes. In Engler & 

Prantl, Die Xatilrlichen Pflanzenfamil- 
ien, ed. 2, 2: 304. Leipzig, 1928. 

(12) . Myxomycetenstudien 16. Die 

Kernphase und die Zahl der Chromo- 
somen. Ber. Deutsche Bot. Ges. 54: 
517-528. 1936. 

(13) Karling, John S. Plasmodiophorales. 

New York, 1942. 

(14) Ledingham, G. A. Zoospore dilation in 

the Plasmodiophorales. Nature 133: 
534. 1934. 

(15) . Occurrence of zoosporangia in 

Spongospora subterranea (Wallroth) 
Lagerheim. Nature 135: 394—395. 

(16) Sixoto, Y., and Yuasa, A. Studies on 

the cytology of reproductive cells. I. On 
the planocytes in five forms of Myxomy- 
cetes. Bot. Mag. (Tokyo) 48: 720-729. 

(17) Smith, E. C. The longevity of myxomycete 

spores. Mvcologia 21: 321-323. 1929. 

(18) . Some phases of spore germina- 
tion of Myxomycetes. Amer. Journ. 
Bot. 16: 645-650. 1929. 

(19) Stosch, H. A. von. Untersuchungen iXber 

die Entwicklungsgeschichte der Myxo- 
mycetes. Sexualitie und Apogamie bei 
Didymiaceen. Planta 23: 623-656. 

(20) Youk, V. Uber den Generationswechsel 

bei Myxomyceten. Oesterr. Bot. 
Zeitsch. 61: 131-139. 1911. [Cited in 
Karling (18).] 

(21) Yuasa, Akira. Studies in cytology of re- 

productive cells. III. The genesis of the 
flagellum in the planocyte of Fuligo sep- 
tica Gmelin. Bot. Mag. (Tokvo) 49: 
538-545. 1935. 

ORNITHOLOGY. — A small collection of birds from Eritrea. 1 Herbert Fried- 
mann, U. S. National Museum. 

During the early stages of World War II 
when North Africa was an important battle- 
field, numbers of American troops were sta- 
tioned in Eritrea, a rather neglected and 
little-known part of eastern Africa. Two of 
the men who were destined to spend some 
time in that former Italian colony collected 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received October 
23, 1947. 

birds as time and opportunity permitted. 
Col. L. R. Wolfe sent in to the IT. S. Na- 
tional Museum a small box containing eight 
birds. A second and larger shipment com- 
prising about 300 specimens was most un- 
fortunately lost in transit. Thane Rine}' 
similarly suffered the loss of the bulk of his 
collection, but was able to bring back some 
37 birds, which he forwarded to the Mu- 
seum of Vertebrate Zoology of the Univer- 



sity of California, whore most of them are 

now. A small number, chiefly duplicates, 
were generously presented by that insti- 
tution to the National Museum, where they 
together with Wolfe's handful of specimens, 
are incorporated with the large East African 
material previously brought together by 
Mearns and others. 

Because of the paucity of published data 
concerning Eritrean birds I thought it ad- 
visable to put on record the contents of the 
Wolfe and Riney collections, small though 
the}' be, and, thanks to the cooperation of 
Dr. Alden H. Miller and Dr. Frank A. Pi- 
telka, I have been able to examine all the 
specimens and to combine them in this re- 
port. A collection containing only 34 species 
could hardly be expected to yield many new 
facts, but as may be seen from the sub j oined 
annotated list, a few items of interest have 
been found to be contained in it. 

Family Accipitridae: Hawks, Eagles, 
and Kites 

Elanus coeruleus coeruleus 


Falco coeruleus Desfontains, Hist, (i.e., Mem.) 
Acad. Roy. Paris, for 1787: 503. 1789 (near 

One specimen, in somewhat abraded plu- 
mage, was collected at Ghinda, altitude 962 
meters, on February 2, 1943, by Thane Riney. 

Melierax metabates metabates Heuglin 

Melierax metabates Heuglin, Ibis 1861: 78 (White 
Nile between 6° and 7° lat. N.). 

During August (16-26), 1942, Wolfe col- 
lected an adult male, adult female, and a Juve- 
nal male 15-20 miles south of Gura. Riney 
obtained an adult (unsexed) between Cheren 
and Agordat, on December 12, 1942. The very 
extensive white freckling on the secondaries 
and inner primaries of the adults suggests that 
they may be somewhat intermediate between 
neumanni and true metabates, but nearer to the 
latter. They have the barred upper tail coverts 
of the nominate race. 

Sclater and Mackworth Praed (Ibis 1919: 
702) consider all Sudanese birds south of Khar- 
toum and north of Lake No as intermediate 
between the two races. Moltoni and Rusconi 
(Gli Uccelli dell' Africa Orientale Italiana 2: 
240. 1942) record neumanni from Eritrea near 

the border of the Red Sea Province of the Su- 
dan, and metabates from the rest of Eritrea. 

The birds are rather small, the adult male 
having a wing length of 300, the female of 307 
mm. In this respect they approach the south- 
west Arabian race ignoscens. 

The August birds show evidences of molting. 

The juvenal female is slightly more rufescent 
on the breast and darker on the upperparts 
than comparable birds from extreme north- 
western Uganda. 

Buteo rofofuscus augur (Ruppell) 

Falco (Buteo) augur Ruppell, Neue Wirbelth., 
Vog.: 38, pi. 16. 1836 (Abyssinia). ' 

A female was taken by Riney about 20 km 
from Decamera, on January 1, 1943, at an ele- 
vation of 2,000 meters. A second specimen is 
unfortunately without data of any kind. Both 
are in the light phase. 

Lophaetus occipitalis (Daudin) 

Falco occipitalis Daudin, Traite 2: 40. 1800 (the 
Anteniquoi country, i.e., Knysna district, Cape 

Wolfe collected a male and an unsexed bird, 
both adults, 15-20 miles south of Gura, August 
16, 1942. He found the crested eagle not un- 
common at elevations of about 5,000 feet. 

Aquila rapax raptor Brehm 

Aquila raptor Brehm, Naumannia 1855: 13 (Blue 
and White Nile). 

An adult female in worn plumage was taken 
by Riney on a nest in a baobab tree, south of 
Barentu, January 17, 1943. The date is in agree- 
ment with Blanford's observation (Geol. and 
Zool. Abyss.: 295-296. 1870) that in Ethiopia 
the birds breed in January. 

Circus pygargus (Linnaeus) 

Falco pygargus Linnaeus, Syst. Nat., ed. 10, 1: 
89. 1758 (Europe). 

Riney collected an unsexed bird (female by 
plumage) 5 km west of Asmara at an elevation 
of 2,409 meters, on January 4, 1943, when he 
saw it foraging over open fields. The bird is in 
rather poor plumage and is molting its remiges, 
making definite identification somewhat diffi- 
cult. The third primary, from the outside, is 
only partly grown in, while the remainder of 
the remiges is considerably abraded. 

Montagu's harrier is a Palearctic bird and 

Apr. 15, 1948 friedmann: a collection of birds from eritrea 


comes into Africa only during the northern 
winter. In eastern Africa it has been recorded 
all the way to South Africa. 

Family Falconidae: Falcons 

Falco naumanni pekinensis Swinhoe 

Falco cenchris var. pekinensis Swinhoe, Proc. 
Zool. Soc. London 1870: 442 (Shihshanling 
(Ming Tombs) near Peking). 

An adult male, taken near Asmara, 2,371 
meters, on January 31, 1943, by Thane Riney, 
is in somewhat worn plumage and differs from 
the nominate race in its darker rufescent back. 
Archer and Godman (Birds of British Somali- 
land and the Gulf of Aden 1: 180-183. 1937.) 
consider this race a rare straggler to their area 
and suggest that it may follow down the Nile 
Valley to the west. If we were to apply this 
suggestion to Eritrea we should have to con- 
sider the bird a rare visitor to that country as 
well, which actually seems to be the case; in 
fact, Riney's specimen is the first record for 
Eritrea, at least as far as published data indi- 
cate. The bird is known to reach South Africa 
during the northern winter and is probably 
commoner along the eastern part of the conti- 
nent than the few records would indicate. 

Falco alopex (Heuglin) 

Tinnunculus alopex Heuglin, Ibis 1861: 69, pi. 3 
(Gallabat, Egyptian Sudan). 

A male, collected by Wolfe, 15 miles south of 
Gura, August 16, 1942, is noticeably darker 
than a female from Talodi, Kordofan, Anglo- 
Egyptian Sudan, the only other example of the 
species available for comparison. Our male also 
differs from this female in having the black 
bars on the median rectrices more complete, 
less marginal in character, and in having those 
on the lateral rectrices less broadened. There 
is a tendency in the female to have these bars 
dilated marginally on all the tail feathers, but 
this is most highly developed on the outer ones. 

Bannerman (Birds Tropical West Africa 1: 
216-219. 1930) has given more extensive com- 
ments on this species than any other recent 
writer. He states that the wing measurement of 
the males varies from 266 to 293 (our example 
measures 276 mm). His account is unfortu- 
nately garbled by some misprinting of his orig- 
inal intention as he then goes on to state that 
the "largest birds are those from N. Nigeria 
with wings d* 212 9 210. ..." 

The intensity (darkness or paleness) of the 
coloration appears to vary individually in this 
kestrel. Bannerman had 23 specimens for study 
and found the darkest birds came from such 
widely separated areas as Kulikoro on the Niger 
River, Jebel Marra in Darfur Province, Anglo- 
Egyptian Sudan, and Ethiopia. On the basis of 
this spotty occurrence of dark birds (among 
which the present Eritrean example appears to 
belong) it seems impracticable to recognize 
Oberholser's race eremica from Togoland, which 
is based entirely on its paler tone. Bannerman 
(loc. cit.) has placed eremica in the synonymy 
of alopex, but, probably by oversight, he uses 
a trinomial for the latter. 

Our specimen shows signs of molt in the 
remiges and rectrices. This suggests the possi- 
bility that its darker tone may be due to the 
freshness of its plumage and that paler birds 
may show the results of fading under the hot 
sun in the dry open country it inhabits. 

Falco tinnunculus tinnunculus Linnaeus 

Falco tinnunculus Linnaeus, Syst. Nat., ed. 10, 1: 
90. 1758, (Europe; restricted type locality, 
Sweden apud Hartert). 

Riney shot a female 10 km from Decamera, 
at 2,050 meters, on January 1, 1943. The bird 
has a wing length of 258 mm and is therefore 
too large to be F. t. archeri Hartert and Neu- 
mann from the Waghar Mountains, and also 
too large to be F. t. carlo (Hartert and Neu- 
mann) of the mountains of East Africa. In 
coloration it is slightly darker than typical 
tinnunculus but not as dark as carlo. It seems 
best treated as a darkish example of the nomi- 
nate race. 

Family Phasianidae: Pheasants, 
Francolins, and Quails 

Francolinus erckelii erckelii (Rtippell) 

Perdix erckelii Rtippell, Neue Wirbelth. Vog.: 12, 
pi. 6. 1835 (Taranta Mountains, Abyssinia). 

Erckel's francolin is a poorly known bird, 
and any additional material of it is still of inter- 
est. Riney collected a female, 40 km north of 
Asmara, 2,390 meters altitude, on January 9, 
1943. It is smaller than any mentioned by Mol- 
toni and Rusconi (Gli Uccelli dell' Africa Orien- 
tale Italiana 3: 22-25. 1944.) having a wing 
length of only 202.6 (their series ranges from 
205 to 230), and a tail length of 91 (as against 
110-140 mm in Moltoni's series). 



Moltoni and Rusconi write that the race pen- 
toni ranges from the Red Sea Province of the 
Sudan into the adjacent parts of Eritrea, but 
as far as I can learn no actual specimens of this 
paler, grayer race; have been taken in Eritrea. 

Family Columbidae: Pigeons and Doves 

Oena capensis capensis (Linnaeus) 

Columba capensis Linnaeus, Syst. Nat., ed. 12, 1: 
286. 1766. (Cape of Good Hope). 

Riney collected a male in low bush near the 
Decamera turn off on the Adi-Ugri Road, 45 km 
south-southwest of Asmara, on January 1, 
1943. The bird is in abraded plumage. 

Family Apodidae: Swifts 

Apus aequatorialis aequatorialis 


Cypselus aequatorialis M tiller, Naumannia 1 : 27. 
1851 (Abyssinia). 

A female mottled swift was taken by Riney 
on the Citao compound at Asmara on February 
2, 1943. It is a bird in rather worn feathering. 

Family Coliidae: Colies, or Mousebirds 

Colius striatus leucotis Rtippell 

Coitus leucotis Rtippell, Mus. Senck. 3: 42, pi. 2. 
1839 (Temben Province, Abyssinia). 

Two unsexed examples of this common spe- 
cies were obtained by Riney, one on the Adi- 
Ugri Road at the base of the Asmara Plateau, 
on November 14, 1942, and one near Cheren, 
on March 15, 1943. This race inhabits Eritrea, 
Bogosland, northern Ethiopia, and adjacent 
portions of the Anglo-Egyptian Sudan. 

Family Coraciidae: Rollers 

Coracias naevius naevius Daudin 

Coracias naevia Daudin, Traite 2: 258. 1800 (Sen- 

"Wolfe obtained one specimen, unsexed, south 
of Gura, on July 26, 1942. It has the white 
stripes on the anterior underparts unusually 
broad and has the purplish brown of the top of 
the head and the greenish of the back slight^ 
darker than in Ethiopian specimens. It has an 
unusually large bill, the culmen measuring 47 
mm from the base ; the largest billed birds from 
Ethiopia and Kenya Colony seen having cul- 
men lengths of 43 mm or less. The specimen 
was molting when collected, the outer remiges 
still showing their sheaths basally. 

Family Cim PIDAE: Hoopoes 

Upupa epops somaliensis Salvin 

Upupa somaliensis Salvin, Cat. Birds Brit. Mus. 
16:13. 1902 (Somaliland). 

On February 2, 1943, Riney obtained an 
adult female on the acacia plain between Ne- 
fasit and Decamara, altitude 1,825 meter-. It 
matches Ethiopian and East Africa examples 
very well. This is the resident race, the nomi- 
nate one being only a winter visitor from Eur- 

Family Bucerotidae: Hornbills 

Tockus nasutus nasutus (Linnaeus) 

Buceros nasutus Linnaeus, Syst. Nat., ed. 12, 1: 
154. 1766 (Senegal). 

The gray hornbill is represented by a head 
collected south of Gura, in August 1942, by 
Colonel Wolfe. 

Tockus erythrorhynchus erythrorhynchus 


Buceros erythrorhynchus Temminck, PI. Col., livr. 
36: sp. 19. 1823 (Senegal). 

Riney collected a male red-billed hornbill on 
the plains west of Agadat, 700 meters altitude, 
on December 13, 1942, and Wolfe shot another 
individual south of Gura, Julj r 26, 1942. 
Riney's bird shows active molt in the tail. 

Family Timaliidae : Babblers 

Turdoides leucopygia leucopygia 

Ixos leucopygius Rtippell, Neue Wirbelth., Yog.: 
82, pi. 30, fig. 1. 1840 (coast of Abyssinia). 

Riney met with this northern race of the 
white-rumped babbler 40 km north of Asmara, 
2,390 meters altitude, on January 9, 1943, when 
he collected a male and an unsexed specimen. 
This race has the whole forehead and fore- 
crown white, while the four more southern sub- 
species have the white reduced or absent. 

The male shows signs of active molt in the 

Family Tttrdidae: Thrushes, Chats, and 

Monticola solitaria solitaria (Linnaeus) 

Turdus solitarius Linnaeus, Syst. Nat., ed. 10: 1: 
170. 1758 (Italy, apud Hartert). 

The blue rock-thrush of central and southern 
Europe winters in northeastern Africa, along 

Apr. 15, 1948 

friedmann: a collection OF BIRDS FROM ERITREA 


with the race longirostris of western Asia. Riney 
collected a female 20 km south-southwest of 
Asmara, 1,930 meters altitude, on January 1, 
1943, which agrees with the nominate form in 
its dark and brownish color (the western Asiatic 
form is paler and more grayish). 

Oenanthe hispanica melanoleuca 


Muscicapa melanoleuca Giildenstadt, Nov. Com. 
Petrop. 19: 468. 1775 (Georgia; Caucasus). 

An unsexed specimen of the eastern black- 
eared wheatear was obtained by Riney 20 km- 
south-southwest of Asmara, 1,930 meters alti- 
tude, on January 1, 1943. The race is known 
to winter from Egypt to Darfur Province in the 
Angol-Egyptian Sudan and to Eritrea and to 
southwestern Arabia (Aden Protectorate). 

Oenanthe lugubris (Riippell) 

Saxicola lugubris Riippell, Neue Wirbelth., Vog.: 
77, pi. 28, fig. 1. 1837 (Simen, Abyssinia). 

The Abyssinian black chat was met with by 
Riney on January 1, 1943, when he collected an 
adult male and female 20 km south-southwest 
of Asmara, at an altitude of 1,930 meters. These 
two specimens are definite evidence that Zed- 
litz (Journ. fur Orn. 1911: 85) was wrong when 
he suggested that this species was migratory in 
Eritrea, leaving for the south after breeding, 
and returning in the latter part of March. To- 
gether with December and January birds pre- 
viously recorded from Ethiopia (Friedmann, 
U. S. Nat. Mus. Bull. 153, pt. 2: 135. 1937.) 
these examples indicate that the migration, if 
any, is very limited in geographical extent. 

Both specimens are in fairly fresh plumage; 
the female has pale tips on the outer rectrices 
(curiously enough, only on the left side of the 
tail, but not on the right), the male has none 
at all. 

Oenanthe isabellina (Temminck) 

Saxicola isabellina Temminck, PI. Col., livr. 79: 
pi. 472, fig. 1. 1829 (Nubia). 

Riney obtained a female isabelline chat in 
the low brushland of the Asmara Plateau, 2,040 
meters elevation, near Asmara, on January 1, 
1943. The species breeds in Europe and east to 
Mongolia and winters in northeastern Africa, 
Arabia, and India. 

Cossypha semirufa semirufa (Riippell) 

Petrocincla semirufa Riippell, Neue Wirbelth., 
Vog.: 81. 1840 (Abyssinia). 

In the dense forest of Monte Marara, 2,340 
meters altitude, about 40 km north of Asmara, 
on January 9, 1943, Riney shot an example (un- 
sexed) of this robin-chat. This must be about as 
far north as it is known to occur; I know of no 
published records north of Bogosland. 

Family Muscicapidae: Old 
World Flycatchers 

Bradornis pallidus bowdleri Collin 
and Hartert 

Bradornis pallidus bowdleri Collin and Hartert, 
Nov. Zool. 34: 52. 1927 (new name for B. p. 
sharpei Rothschild, 1913, not B. sharpei Bo- 
cage, 1894: Abyssinia). 

One female was collected by Riney near De- 
camera, 2,000 meters altitude, on January 1, 
1943. This race of the pale flycatcher inhabits 
northern Ethiopia (south to the vicinity of Adis 
Ababa) and Eritrea, where it lives on open 
bushy areas. 

Batis minor erlangeri Neumann 

Batis minor erlangeri Neumann, Journ. fur Orn. 
1907: 352 (Gara Mulata, near Harrar, Ethio- 

Riney obtained an unsexed specimen (male 
by plumage characters) near Decamera, 2,000 
meters altitude, on January 1, 1943. Inasmuch 
as this example agrees with Ethiopian speci- 
mens of erlangeri and shows no approach to the 
characters ascribed to chadensis, I have no 
hesitancy in so classifying it, although it ex- 
tends the known range of erlangeri northward 
a very considerable distance. It was known pre- 
viously from the Harrar area in central eastern 
Ethiopia, southwest to southern Shoa, and to 
Lake Stefanie. 

Family Motacillidae : Wagtails and Pipits 

Anthus richardi cinnamomeus Riippell 

Anthus cinnamomeus Riippell, Neue Wirbelth., 
Vog., 103. 1840 (Simien Province, Abyssinia). 

On January 1, 1943, Riney collected a female 
of this pipit about 15 km south-southwest of 
Asmara, at 2,040 meters altitude. The bird was 
seen on the ground in low brushlands. The 
specimen, which is in somewhat frayed plumage 
agrees well with others from Ethiopia. 



Family Laniidae: Shrikes 

Lanius collaris humeralis Stanley 

Lanius humeralis Stanley, in Salt, Travels in 
Abyssinia . . . , Appendix, li, no. 4. 1814 (Cheli- 
cut, Abyssinia). 

One female, collected by Riney 15 km south 
southwest of Asmara, 2,040 meters, January 1, 
1943, is of this subspecies, which occurs from 
Eritrea and Ethiopia, south through eastern 
Africa (west to central Uganda), to Zululand 
and Natal. 

Family Prionopidae: Wood-shrikes 

Prionops cristata cristata Ruppell 

Prionops (Lanius) cristatus Ruppell, N. Wir- 
belth., Vog., lief. 183: 30, pi. 12, fig. 2. 1837 
(coast at Massawa). 

Riney obtained an unsexed bird at Ghinda, 
962 meters altitude, on February 2, 1943. The 
specimen is in very worn feathering. 

Family Sturnidae : Starlings 

Lamprocolius chalybeus chalybeus 

(Hemprich and Ehrenberg) 

Lamprotornis chalybeus Hemprich and Ehrenberg, 
Symbolae physicae, folio y: pi. 10. 1828 (Ara- 
bukol, Dongola). 

A female, showing evidence of active molting 
in the wings, was collected by Riney 30 km 
from Asmara, 1,930 meters altitude, on Janu- 
ary 1, 1943. 

Family Ploceidae: Weaverbirds 

Sporopipes frontalis abyssinicus Mearns 

Sporopipes frontalis abyssinicus Mearns, Smith- 
sonian Misc. Coll. 56(14): 7. 1910 (Abyssinia). 

Three examples of the speckle-fronted 
weaver were collected by Riney, 2 males and 
1 female, on the acacia plain between Nefasit 
and Decamera, elevation 1,852 meters, Feb- 
ruary 7, 1943. One of the males shows evidence 
of molting in the tail. 

These specimens are somewhat darker on the 
upper surface of the wings than is the type, 
but this may be due to the fact that the latter 
is in very fresh plumage and has more exten- 
sive pale margins to these feathers. 

Ploceus baglafecht baglafecht (Daudin) 

Loxia baglafecht Daudin, in Buffon, Hist. Nat. 
(Didot's ed.), Quadrupeds 14: 245. 1799 (ac- 
tually 1802) (Abyssinia). 

Riney collected a male, 40 km north of As- 
mara, 2,390 meters elevation, on January 9, 
1943. This weaver is known to occur at alti- 
tudes of from 5,000 to 12,000 feet in Bogosland 
and Ethiopia. 

Uraeginthus bengalus bengalus 


Fringilla bengala Linnaeus, Syst. Nat., ed. 12, 1: 
323. 1766 ("Bengal"; Senegal substituted by 
Sclater, Syst. Avium P^thiop. 2: 804. 1930). 

One male and one female were taken by 
Riney on January 1, 1943 in low acacia-dotted 
open grassy country between Decamera and the 
Adi-Ugri Road, 2,000 meters elevation. The 
male had the testes enlarged. Both birds are in 
worn plumage. 

I can see little advantage in accepting the 
recent suggestion that bengalus and its races 
are conspecific with the angolensis group. There 
is more to be said for Delacour's action in 
"lumping" Uraeginthus in the genus Estrilda, 
but even in this I hesitate to follow him as the 
cordon-bleus are a very distinctive section, at 
least, of the waxbill aggregate. 

Family Fringillidae : Finches, Sparrows, 
and Buntings 

Poliospiza tristriata tristriata 


Serinus tristriatus Ruppell, Neue Wirbelth., Vog.: 
97, pi. 35, fig. 2. 1840 (Taranta Pass, Abys- 

Riney collected a female in the Bermuda 
grass at the base of the Asmara Plateau, at 
1,930 meters elevation 20 km south-southwest 
of Asmara, January 1, 1943. This seedeater ap- 
pears to range from altitudes of from 4,000 to 
11,000 feet, and is reported (in literature)" to be 
common in Eritrea and northern Ethiopia. 

Fringillaria tahapisi septemstriata 

Emberiza septemstriata Ruppell, Neue Wirbelth., 
Vog.: 86, pi. 30. 1840 (Gondar, Abyssinia). 

A male in fairly worn plumage was taken by 
Riney on January 1, 1943, 45 km from Asmara, 
1,930 meters elevation. 

This race of this rock bunting is characterized 
by the extensive rufescent area on the basal 
portion of the inner web of the outermost pri- 

Apr. 15, 1948 



ZOOLOGY. — Some echinoderms from Biak, Schouten Islands. 1 
and Frederick M. Bayer, U. S. National Museum. 

Austin H. Clark 

Biak (or Wiak) is the largest and most 
easterly of the Schouten Islands, a small 
group of islands just north of Geelvink Bay, 
New Guinea, in approximately latitude 1°S. 
It is 45 miles long and 23 miles wide. It is a 
low island, not rising more than 50 feet 
above sea level except at the southern end 
where there is a hill 800 feet in height. 

Until the late war Biak was regarded as 
of little importance. It was remote from the 
Netherlands East Indies capital of Batavia, 
and the natives, addicted to headhunting 
and other unpleasant practices, were hostile 
to strangers. Few collections of any kind 
had ever been made there, and no echino- 
derms have ever been recorded from the 

During the war the junior author, then a 
member of a 5th Airforce Photo Recon- 
naisance Squadron, was fortunate enough 
to spend four months at Sarido village 
within a stone's throw of a luxuriant coral 
reef. This reef is a fringing reef of rather in- 
significant proportions. It is perhaps 200 
yards wide at the most and approaches 
quite close to the shore, which is sandy in 
the small embayment in which Sarido 
stands, and rocky southward toward the 
point on which the airstrip is located. In the 
little bay the actively growing reef is some- 
what farther from shore than elsewhere and 
is separated from it by a strip of slightly 
deeper water, from 3 to 5 or 6 feet deep at 
low tide, with a sandy bottom on which are 
scattered living and dead clumps of coral 
and patches of eelgrass and algae. This is a 
rather restricted zone, grading rapidly into 
the active reef zone. 

The reef itself is covered with water at all 
but the very lowest tides when a few of the 
highest coral prominences are exposed. It is 
made up of great masses of living coral, 
huge rounded heads of massive species 
many feet thick and submarine "brier 
patches" of branched madrepores in which 
reef fishes of indescribable colors hide by 
day. It is not an algal reef of the type found 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received November 
7, 1947. 

in the Marshall Islands and generally 
throughout the Pacific. There is no promi- 
nent ridge of Lithothamnion, nor are these 
plants even present in noticeable numbers. 

The coral masses afford precarious foot- 
holds, for the delicate Acroporas and other 
branched species are liable to give way, 
plunging one's foot into 4 or 5 feet of water 
and raking one's legs with razor-sharp edges 
in the descent. Among these coral masses 
are pools of deep water with sandy bottoms, 
providing a fine habitat for marine ani- 

Two invertebrates especially force them- 
selves on the attention of the collector, the 
giant clam, Tridacna, with its mantle rich 
shades of blue, green, and purple, and a 
starfish, Linckia laevigata, with a very small 
disk and long, rigid, cylindrical arms of the 
most outlandish blue imaginable. These 
two creatures are everywhere. Coral masses 
are studded with the multicolored zigzags 
formed by the clams, and the entire reef is 
dotted here and there on sand and coral 
alike with the brilliant blue 5-pointed stars. 

The most interesting animals were less 
evident, and it required poking into dark 
crevices, turning over coral blocks, and stir- 
ring up sand pockets to find them. Occa- 
sionally one of the giant slate-pencil urchins, 
Heterocentrotus mammillatus, would be 
found in the open, though usually they were 
tucked away in remote nooks and cran- 
nies, their presence betrayed only by a stray 
spine or two projecting from a small open- 
ing, apparently much too small to allow the 
creature free passage. Cidarids could be 
found in similiar situations, though they 
were much less common. 

Turning over coral heads was a simple 
way of finding interesting things. Brittle- 
stars were, of course, under every one, and 
those with large enough recesses often con- 
tained comatulids of unusual beauty. The 
less conspicuous starfishes also were often 
found in such situations. The little sand 
pockets under the rocks produced very in- 
teresting mollusks, such as the venomous 
cone shells, Conus textile, C. striatus, and C. 
geographus, not to mention dozens of less 
conspicuous things such as abalones (Halio- 



ta's), several species of Trochus, Mitra, 
Cymatium, and many other genera. 

The large starfishes Culcita novaeguincae 
and Acanthaster planci were found crawling 
about in the open, but because of their pro- 
tective coloration were more often over- 
looked than seen. The latter is adorned with 
the most vicious 3-cornered spines imagin- 
able. In collecting one of these the spines 
drew blood through a pair of heavy leather 

Toward the outer reef the water deepens, 
the coral heads become more widely spaced, 
and, reaching the brink, one can look down 
the almost vertical wall into fathomless 
blue. Sharks were sometimes seen cruising 
along in the hazy middle distance, and al- 
wsljs myriads of reef fishes, moorish idols, 
parrotfishes, pomacentrids, wrasses, scor- 
pionfishes, and many others. To the right 
and left on almost every coral pinnacle 
could be seen a huge black or deep red 
comatulid, gently swaying in the aquatic 
breezes. These seemed never to stray from 
their own chosen perch, for on several visits 
to the same spot we found them unchanged. 

The occurrence of Heterocentrotus mam- 
millatus at Biak is especially interesting, for 
only H. trigonarius has been definitely re- 
corded from nearby New Guinea. Hetero- 
centrotus trigonarius occurs in the Philip- 
pines and generally throughout the Pacific 
islands. It is replaced by H. mammillatus in 
the Hawaiian, Bonin, and Riu Kiu islands, 
which is also found at Lord Howe Island, 
in the Murray Islands at the northern end 
of the Great Barrier reef, and at Cape Jau- 
bert, Western Australia. 

The specimens listed below are in the U. S. 
National Museum, and the numbers follow- 
ing the names are those in the catalogue of 
the Division of Echinoderms. 


Comantheria briar eus (Bell), 4, E.6935, E.6937, 

E.6939, E.6961. 
Comanthus bennetti (J. Mtiller), 4, E.6963, 

E.6970, E.6972, E.6976. 
Comanthus timorensis (J. Muller), 5, E.6936, 

E.6954, E.6962, E.6874. 

Comanthus parvicirra (J. Muller), 5, E.G841, 

B.6842, E.G9GG, E.6971. 
Himerometra magnipinna (A. H. Clark), 7, 

E.6957, E.G958, E.6964, E.6965, E.6973. 
Stephanometra spicata (P. H. Carpenter), 1, 

Stephanometra protectus (Liitken), 1, E.6843. 
Lamprometra palmata palmata (J. Muller), 4, 

E.6959, E.6960, E.6969, E.G987. 


Plococidaris verticillata (Lamarck), 1, E.6967. 
Eucidaris metularia (Lamarck), 3, E.6968. 
Mespilia globulus (Linne), 1, E.6951. 
Heterocentrotus mammillatus (Linn6), 1, E.G977. 


Ar chaster typicus Muller and Troschel, 1, 

Protoreaster nodosus (Linne), 1, E.6830. 
Culcita novaeguineae Muller and Troschel, 3, 

Gomophia aegyptica Gray, 1, E.6828. 
Nardoa mollis de Loriol, 2, E.6981. 
Linckia laevigata (Linne), 4, E.6982-E.6984. 
Linckia multifora (Lamarck), 3, E.6986. 
Linckia guildingii Gray, 1, E.6985. 
Asterope carinifera (Gray), 1, E.6827. 
Asterina cepheus (Muller and Troschel), 1, 

Othilia luzonica Gray, 1, E.6978. 
Acanthaster planci (Linne), 2, E.6955, E.6956. 


Ophiodera brevispina (von Martens), 1, E.6950. 

Ophiothrix longipeda (Lamarck), 1, E.6944. 

Ophiocoma erinaceus (Muller and Troschel), 1, 

Ophiocoma scolopendrina (Lamarck), 2, E.6953. 

Ophiomastix annulosa (Lamarck), 2, E.6949. 

Ophiomastix liltkenii Pfeffer, 2, E.6948. 

Ophiarthrum pictum (Muller and Troschel), 5, 
E.6946, E.6947. 

Ophiarachna incrassata (Lamarck), 1, E.6943. 

Ophiarachnella septemspinosa (Muller and Tro- 
schel), 1, E.6945. 

Ophiolepis superba H. L. Clark, 5, E.6940. 

Ophiolepis cincta Muller and Troschel, 2, 

Ophioplocus imbricatus (Muller and Troschel), 
5, E.6941. 

Officers of the Washington Academy of Sciences 

President Frederick D. Rossini, National Bureau of Standards 

Secretary C. Lewis Gazin, U. S. National Museum 

Treasurer Howard S. Rappleye, Coast and Geodetic Survey 

Archivist Nathan R. Smith, Plant Industry Station 

Custodian and Subscription Manager of Publications 

Harald A. Rehder, U. S. National Museum 

Vice-Presidents Representing the Affiliated Societies: 

Philosophical Society of Washington Walter Ramberg 

Anthropological Society of Washington T. Dale Stewart 

Biological Society of Washington John W. Aldrich 

Chemical Society of Washington Charles E. White 

Entomological Society of Washington C. F. W. Muesebeck 

National Geographic Society Alexander Wetmore 

Geological Society of Washington William W. Rubey 

Medical Society of the District of Columbia Frederick O. Coe 

Columbia Historical Society Gilbert Grosvenor 

Botanical Society of Washington Ronald Bamford 

Washington Section, Society of American Foresters William A. Dayton 

Washington Society of Engineers Clifford A. Betts 

Washington Section, American Institute of Electrical Engineers 

Francis B. Silsbee 

Washington Section, American Society of Mechanical Engineers 

Martin A. Mason 

Helminthological Society of Washington Aurel 0. Foster 

Washington Branch, Society of American Bacteriologists Lore A. Rogers 

Washington Post, Society of American Military Engineers . Clement L. Garner 

Washington Section, Institute of Radio Engineers Herbert Grove Dorsey 

Washington Section, American Society of Civil Engineers Owen B. French 

Elected Members of the Board of Managers: 

To January 1949 Max A. McCall, Waldo L. Schmitt 

To January 1950 F. G. Brick wedde, William W. Diehl 

To January 1951 Francis M. Defandorf, William N. Fenton 

Board of Managers All the above officers plus the Senior Editor 

Board of Editors and Associate Editors [See front cover] 

Executive Committee Frederick D. Rossini (chairman), Walter Ramberg, 

Waldo L. Schmitt, Howard S. Rappleye, C. Lewis Gazin 

Committee on Membership 

Harold E. McComb (chairman), Lewis W. Butz, C. Wythe Cooke, William 

W. Diehl, Lloyd D. Felton, Regina Flannery, George G. Manov 

Committee on Meetings'. Raymond J. Seeger (chairman), 

Frank P. Cullinan, Fred L. Mohler, Francis O. Rice, Frank Thone 

Committee on Monographs: 

To January 1949 Lewis V. Judson (chairman), Edward A. Chapin 

To January 1950 Roland W. Brown, Harald A. Rehder 

To January 1951 William N. Fenton, Emmett W. Price 

Committee on Awards for Scientific Achievement (Karl F. Herzfeld, general chairman): 

For the Biological Sciences 

C. F. W. Muesebeck (chairman), Harry S. Bernton, Chester W. Emmons, 
Elmer Higgins, Mario Mollari, Gotthold Steiner, L. Edwin Yocum 

For the Engineering Sciences 

Harry Diamond (chairman), Lloyd V. Berkner, Robert C. Duncan, 
Herbert N. Eaton, Arno C. Fieldner, Frank B. Scheetz, W. D. Sutcliffe 

For the Physical Sciences 

Karl F. Herzfeld (chairman), Nathan L. Drake, Lloyd D. Felton, 
Herbert Insley, William J. Rooney, Robert Simha, Michael X. Sullivan 

Committee on Grants-in-aid for Research 

. .F. H. H. Roberts, Jr. (chairman), Anna E. Jenkins, J. Leon Shereshevsky 

Representative on Council of A. A. A. S Frank Thone 

Committee of Auditors 

William G. Brombacher (chairman), Harold F. Stimson, Herbert L. Haller 

Committee of Tellers 

...John W. McBurney (chairman), Roger G. Bates, William A. Wildhack 



Biochemistry. — The chemical nature of enzymes. James Batchel- 

ler Sumner 113 

Chemistry. — Dr. Stephen Brunauer's contributions in the field of 

adsorption. Ralph A. Beebe 117 

Meteorology. — Loose usage of weather words. W. J. Humphreys. . 123 

Paleontology. — An interesting occurrence of fossil tracks in West 

Virginia. David H. Dunkle 130 

Mycology. — The swarm-cells of Myxomycetes. Eugene W. Elliott 133 

Ornithology. — A small collection of birds from Eritrea. Herbert 

Friedmann 137 

Zoology. — Some echinoderms from Biak, Schouten Islands. Austin 

H. Clark and Frederick M. Bayer 143 

This Journal is Indexed in the International Index to Periodicals 


Vol. 38 

May 15, 1948 

No. 5 




Jambs I. Hoffman 


Alan Stone 


Frank C. Kracek. 



Lawrence A. Wood 


J. P. E. Morrison 


Elbert L. Little, Jr. 


Richard E. Blackwelder 


James S. Williams 


Waldo R. Wedel 


Irl C. Schoonover 




450 Ahnaip St. 

at Menasha, Wisconsin 


Entered as second class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a speoial rate of postage provided for in the Aot of February 28, 1925. 

Authorized January 21, 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In orderjjjto facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author* 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate): 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 .50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.).. .-. . . 8.00 0.60 

Vol. 23 to current vol. (12 numbers per vol.) 7 .50 .90 

• Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 




Vol. 38 

May 15, 1948 

No. 5 

PHYSICS. — General survey of certain results in the field of high pressure physics. 1 
Percy W. Bridgman, Harvard University. 

In this lecture I shall attempt to present 
a general survey of those parts of the field 
of high pressure physics with which I have 
had direct contact, dealing first with tech- 
nical matters of producing and measuring 
high pressure, and secondly with the physi- 
cal phenomena which occur under high 

With regard to technique, several differ- 
ent ranges of pressure are to be recognized. 
The first step was to devise a method of 
packing which should be without leak, 
since leak had limited the range of previous 
experiments. A packing was devised, shown 
in Fig. 1, which automatically becomes 
tighter the higher the pressure, so that any 
pressure is accessible up to the strength of 
the containing vessels. If the vessels are 
made of one-piece construction, from the 
best heat treated alloy steels, it is possible 
to reach pressures of 12,000 kg/cm 2 as a 
routine matter and on occasion for short 
intervals of time as high as 20,000. For 
many years my work was confined to this 
range, and in this range it proved feasible 
to measure nearly all the ordinary physical 
properties of substances. The next step was 
to give the pressure vessel external support 
which increases in magnitude at the same 
time the internal pressure increases. A sim- 
ple method of doing this is to make the ex- 
ternal surface of the pressure vessel conical 
in shape, and to push it into a heavy collar 
with a force which increases as the internal 
pressure increases, as illustrated in Fig. 2. 
With apparatus of this kind it is possible to 
make routine experiments up to 30,000 

1 Nobel Laureate Lecture delivered at Stock- 
holm, Sweden, December 11, 1946. Reprinted by- 
permission from the proof of the article for Les 
Prix Nobel en H946. 

kg/cm 2 with volumes of the order of 15 cm 3 , 
to get electrically insulated leads into the 
apparatus, and practically to repeat all the 
former work in the range to 12,000. I am 
still engaged in carrying out this program. 
An extension of the same technique on a 

Fig. 1. — The general scheme of the packing by 
which pressure in the soft packing materials is 
automatically maintained a fixed percentage 
higher than in the liquid. 

smaller scale with capacities of the order of 
0.5 cm 3 can be made up to 50,000 kg/cm 2 . 
In this range all ordinary liquids freeze 
solid, electrically insulated leads cannot be 
got into the apparatus, and the phenomena 
which can be studied are limited to various 
volume effects, such as compressibilities and 
phase changes, including fusions and poly- 
morphic transitions. 




The external support of the vessel is only 
one of the factors that make possible the 
extension of range from 12,000 to 50,000. 
No steel piston will support as much as 
50,000; carboloy, however, the recently de- 
veloped substance for tools formed by 
cementing a fine powder of tungsten carbide 
with cobalt, fortunately proves to have a 
compressive strength high enough for the 

Fig. 2. — Illustrating the general principle of the 
method for giving external support to the pressure 
vessel in such a way that support increases auto- 
matically with the increase of internal pressure. 

The next step in extension of range, from 
50,000 to 100,000 kg/cm 2 , demands still 
more effective support of the pressure ves- 
sel. This is done by immersing the entire 
pressure vessel in a fluid under pressures 
ranging up to 30,000 kg/cm 2 . The pressure 
apparatus has to be made still smaller, the 
pistons are only 1.6 mm in diameter, and 
the capacity is only a few cubic millimeters. 
The pressure cylinder itself, as well as the 
pistons, is now made of carboloy with an 
external jacket of shrunk-on steel to give it 
greater strength. The piezometer is illus- 
trated in Fig. 3. Even with this type of con- 
struction so great an extension of range as 
from 50,000 to 100,000 would not have been 
possible if it were not for a fortunate change 
in the properties of metals under pressure. 
At pressures of 25,000 kg/cm 2 ordinary 
grades of steel become capable of almost 
indefinite deformation without fracture, so 
greatly has their ductility been increased, 
as shown in Fig. 4. Even carboloy loses its 
normal brittleness and becomes capable of 
supporting higher tensile stresses without 
fracture than steel. 

Up to the present, the compressibilities 
and polymorphic transitions of some 30 ele- 
ments and simple compounds have been 

studied in the range to 100,000 kg/cm 2 . 

Much higher pressures than 100,000 can 
be reached in very small regions by con- 
structing the apparatus entirely of carboloy, 
but up to the present no particularly im- 
portant physical results have been attained 
in this range. 

In addition to the problem of attaining 
the pressures, there is the problem of meas- 
uring them and measuring the effects which 
they produce. This demands in the first 
place the establishment of various fixed 
points. In the range up to 30,000 a sufficient 
number of such points has been established 
to permit measurements to an accuracy of 
about 0.1 percent. A transition of bismuth 
in the neighborhood of 25,000 gives one 
convenient such point. An essential part of 
the measuring technique is the utilization of 
the change of resistance of manganin under 
pressure, first suggested by Lisell at 
Uppsala. Above 30,000 the territory is not 
so well marked out; it is probable that the 
measurements to 100,000 have an accuracy 
of about 2 percent. 






Fig. 3. — The miniature apparatus 
for reaching 100,000 kg /cm 2 . 

It is natural to think of volume compres- 
sion as the simplest and most fundamental 
of all the effects of hydrostatic pressure, and 
for that reason it will be discussed first here. 
It is not, however, the simplest to measure 
experimentally, because the measurements 
immediately obtained are relative to the 
containing vessel, which is itself distorted. 

May 15, 1948 

bridgman: the field of high pressure physics 


Elaborate procedures may be necessary to 
eliminate the effect of such distortion. 

The compression of gases is outside the 
range of this work; at pressures of 1,000 
kg/cm 2 or more the densities of gases be- 
come of the same order of magnitude as 
those of their liquid phase, and there ceases 
to be any essential difference between gas 
and liquid. If the volume of any ordinary 
liquid is plotted as a function of pressure at 
constant temperature, a curve will be ob- 
tained which at low pressures has a high 
degree of curvature and a steep tangent, 
meaning a high compressibility, but as pres- 
sure increases the curvature rapidly be- 
comes less and the curve flattens off. In 
Fig. 5 the volume of a typical liquid, ether, 
is shown as a function of pressure. For com- 
parison, the curve of the most compressible 
solid, caesium, is also shown. Two different 
physical mechanisms are primarily responsi- 
ble for the different behavior in the low and 
high pressure ranges. The low range of high 
compressibility is the range in which the 
chief effect of pressure is to push the mole- 

cules into closer contact, eliminating the 
free spaces between them. In this range 
individual substances may show large and 
characteristic individual differences. In the 
higher range the molecules have been 
pushed into effective contact, and the com- 
pressibility now arises from the decrease of 
volume of the molecules themselves. This 
effect persists with comparatively little 
decrease over a wide range of pressure. This 
effect is of course present also in the lower 
range of pressure, but there it is masked by 
the much larger effect arising from squeez- 
ing out the free spaces between the mole- 
cules. If one attempts to set up a formula 
for the effect of pressure on volume on the 
basis of measurements in the low range only, 
one will be likely to neglect too much the 
contribution from the compressibility of the 
molecules, with the result that the actual 
volumes at high pressures will be found to 
be materially smaller than the volumes 
which would be extrapolated from the low 
pressure formulas. This, as a matter of fact, 
has been a property of practically all the 

Fig. 4. — Illustrating the effect of pressure in increasing the ductility of steel. On the left, a piece 
of mild steel broken in tension at atmospheric pressure. On the right, the same steel pulled to a much 
greater reduction of area without fracture in a liquid at 25,000 kg/cm 2 . 



VOL. 38, NO. 5 

formulas that have been derived from low 
pressure data. 

At high pressures, the volumes of ordi- 
nary organic liquids become surprisingly 
alike in spite of initial differences. To illus- 
trate the rapid falling off of compressibility 
with pressure, the volume change in the 
first 5,000 kg/cm 2 is roughly the same on the 
average as the volume change between 
5,000 and 50,000; the effect is accentuated 
by the fact that the volume decrement in 
the latter range often includes the volume 
discontinuity on freezing. 

In the low pressure range, in which the 
molecules are being pushed into effective 
contact, one might expect effects depending 
on the shapes of the molecules, and that 
these effects would be highly specific with 
the liquid. This is the case. In the low pres- 
sure range a great variety of small-scale ab- 
normalities are superposed on the larger 
scale uniformities, and these small-scale 
effects vary greatly from liquid to liquid. 
Thus there may be sub-ranges of an extent 
of a few thousand kg/cm 2 in which the com- 
pressibility increases with increasing pres- 
sure instead of decreasing as is normal, or 

the thermal expansion may also increase 
with increasing pressure instead of decreas- 
ing. Any satisfactory theory of liquids must 
ultimately give an account of these small- 
scale effects but for the present the large- 
scale effects must have first attention. When 
the theory of liquids does come to be writ- 
ten, the first step may well be to set up an 
idealized "perfect liquid" in analogy to the 
perfect gas which has played so important a 
role in the theory of gases. The experimen- 
tal results at high pressures show sufficient 
uniformity in the behavior of all ordinary 
organic liquids to indicate that such an 
idealized perfect liquid is not too far from 
the actuality. 

The compressibility of solids varies over 
a much wider range than does that of the 
ordinary liquids; caesium, for example, is 
350 times more compressible than diamond. 
The highest compressibilities among solids, 
judging from indirect evidence, are probably 
to be found in solid hydrogen and helium. 
As in the case of liquids, the compressibility 
of solids normally drops off with increasing 
pressure. This would be expected in general 
because of the operation of a law of "di- 














I 2 3 4 5 6 7 8 


Fig. 5. — Volume as a function of pressure for a typical liquid, ether. The corresponding curve is also 
shown for caesium, the most compressible solid. The liquid is initially much more compressible than 
the solid, but at higher pressures is less compressible. 

May 15, 1948 bridgman: the field of high pressure physics 


minishing returns," and is obviously neces- 
sary when pressure is raised indefinitely 
because if volume continued to decrease at 
its initial rate it would eventually become 
negative. For instance, the volume of 
caesium would become negative at a pres- 
sure of only 14,000 kg/cm 2 if it contin- 
ued to decrease with pressure at the ini- 
tial rate. In spite of the fact that the 
compressibility of solids on the average 
must decrease with increasing pressure, 
there is a very marked qualitative dif- 
ference as compared with liquids. The 
initial phase of very rapid decrease is ab- 
sent, and the decrease is spread more uni- 
formly over the entire pressure range. The 
difference is to be accounted for by the lat- 
tice structure of solids; with increasing 
pressure the atoms retain their position in 
the lattice with the result that a smaller 
part of the free space between the atoms is 
available for occupancy as the centers of 
the atoms are forced closer together. 

The volume decrements of a number of 
the more compressible solids are shown as a 
function of pressure up to 100,000 kg/cm 2 
in Fig. 6. The curvature is in general very 

There is no thermodynamic necessity 
that the compressibility should decrease 
with increasing pressure, although this 
opinion has sometimes been expressed. Solid 
substances are known in which the com- 
pressibility may increase with increasing 
pressure over a comparatively wide range of 
pressure. The most striking example is 
quartz glass. The compressibility not only 
increases with pressure, but increases at an 
accelerating rate. This continues up to 
35,000 kg/cm 2 and then abruptly stops. At 
this pressure there is a discontinuity in the 
derivative, a transition of the "second kind" 
in the nomenclature of Ehrenfest, and from 
here on compressibility decreases with rising 
pressure as is normal. The mechanism 
which is responsible for the low pressure 
effect abruptly ceases to act. Fig. 7 shows 
the relations. 

So far we have been considering the effect 
of pressure on the volume of isotropic sub- 
stances; this includes substances like glass 
and all cubic crystals. If the material crys- 
tallizes in some non-cubic system, the ef- 

fects are more complicated. The compressi- 
bility is not the same in all directions, so 
that the shape of bodies composed of such 
crystals may change under pressure. The 
differences of compressibility in different di- 
rections may be large; thus zinc is eight 
times as compressible in the direction of the 
hexagonal axis as at right angles to it. Some 
difference in this direction might be ex- 
pected, because the atomic spacing is 
greater along the axis than at right angles, 
but no simple consideration would lead to 
the expectation of differences as large as 
this. There is even one substance, tellurium, 
which has a negative compressibility along 
the axis. That is, when a single crystal of 
tellurium is subjected to hydrostatic pres- 
sure by a fluid in which it is completely im- 
mersed, it expands along the axis. 

Considerable success has been achieved 
in calculating theoretically the effect of 



Fig. 6.— The volume compression of several 
elements up to 100,000 kg/cm 2 . The breaks in 
some of the curves indicate polymorphic tran- 



pressure on the volume of simple solids. The 
first success was with simple ionic lattices of 
the type of NaCl by Max Born, who was 
able to get acceptable values for the lattice 
spacing and for the initial compressibility. 
He was not at first successful, however, in 
reproducing the change of compressibility 
with pressure, and even today complete 
success has not been attained in this regard. 
More complete results have recently been 
obtained for the alkali metals by applying 
the methods of wave mechanics. Bardeen 
has had surprising success in reproducing 
the entire volume curve over the experi- 
mental pressure range for the alkali metals. 
The calculations are particularly simple 
here because there is only one free electron 
per atom, and it turns out that the major 
part of the effect arises from the increase of 
kinetic energy of the free electrons arising 
from their decrease in effective wave length 
when the volume is decreased. Other metals, 
with more free electrons, are more difficult 
to compute, but it is anticipated that the 
difficulties are merely difficulties of the com- 
plexity of the calculation. 

Theory is apparently not yet in a position 
to attack with much success the problem of 
non-cubic crystals. 

We now consider the discontinuous vol- 
ume effects arising from phase changes of 
various sorts. The simplest of these is the 
effect of pressure on melting. Historically 
the study of the effect of pressure on melt- 







ing was approached with the anticipation 
that the effects would be found to be similar 
to the effect of pressure on vaporization, 
and in particular that there would be criti- 
cal phenomena, so that above a certain 
pressure and temperature continuous pas- 
sage would be possible between liquid and 
solid. It soon appeared, however, that the 
pressure scale of any such effects must be 
much more extensive than the scale of the 
critical effects between liquid and vapor, 
and that whereas pressures of a few hun- 
dred kg/cm 2 were adequate in the latter 
case, pressures of thousands of kg/cm 2 
would be required to produce analogous ef- 
fects for solid and liquid, if indeed they 
could be produced at all. With every exten- 
sion of pressure range the probability of the 
existence of any such critical phenomena 
has become increasingly remote. Melting 
curves have now been followed up to 40,000 
kg/cm 2 ; a number of these are shown in 
Fig. 8. The melting curves of all substances 
have certain qualitative features in com- 
mon, so that it is appropriate to speak of 
"the" melting curve just as one may speak 
of "the" vaporization curve. In other re- 
spects, however, the situation with regard 
to melting is qualitatively different from 
that with regard to vaporization. In par- 
ticular, all melting curves, that is, the curve 
of melting temperature against pressure, 
are concave toward the pressure axis with a 
curvature becoming less at higher pressures, 





Pressure Kg/cm a 

SiO* Glass 

Fig. 7. — The deviations from linearity of the volume decrements of quartz glass for pressure in- 
crements of 5,000 kg/cm 2 plotted against pressure. The cusp in the curve marks the change from_ab- 
normal to normal behavior. 

May 15, 1948 



and the curve of difference of volume be- 
tween liquid and solid as a function of 
pressure is convex toward the pressure axis 
with a curvature decreasing with increasing 
pressure. No critical point has ever been ob- 
served in the experimental range. If there 
were such a point outside the range, the 
latent heats and the volume difference be- 
tween liquid and solid would have to vanish 
at a common pressure and temperature. 
Extrapolation of the curves for latent heat 
and volume difference indicates that neither 
of them will vanish at any finite pressure or 
temperature, to say nothing of both vanish- 
ing at the same pressure and temperature. 
The probability at present seems over- 
whelming that there can be no critical point 
between liquid and solid, at least for the 
type of substance investigated up to now, 
which includes organic substances of vari- 
ous types and a few metals. The same line 
of argument rules out the existence of other 
such features on the melting curve as a 
maximum temperature or an asymptotic 
temperature. In general, the melting curve 
rises to indefinitely high temperatures with 
indefinitely increasing pressure but at a 
diminishing rate, the curve becoming more 
nearly linear. 

It is possible to show thermodynamically 
that if a substance expands when it melts, 
its melting temperature must rise with in- 
creasing pressure, and, conversely, it falls. 
There are only three substances which be- 
long to the latter category in the ordinary 
range, water, bismuth, and gallium. Con- 
sistent with thermodynamics the melting 
curves of these three substances are found 
to fall. Furthermore, the curvature in- 
creases and the curves drop more and more 
rapidly as pressure increases. Such a state 
of affairs apparently cannot continue in- 
definitely. Nature extricates itself from the 
dilemma by the "liquidating" of such ab- 
normal substances. Above a certain pres- 
sure the lattices in which these substances 
initially crystallize become unstable, and 
the lattice collapses into another lattice. 
The new lattice has a volume so much less 
than the former lattice that the solid phase 
is now more dense than the liquid, and from 
here on the melting curve rises as for other 
substances. The collapse of the lattice oc- 

10400 ZO.000 

Rressure , K<j )«m 2 


Fig. 8. — Melting temperature against pressure 
for a number of substances. At 15,000 kg/cm 2 
the order of substances, reading from top down, 
is chloroform, chlorobenzene, chlorobenzene 
(second modification), water (ice VI), n-butyl 
alcohol, carbon bisulphide, methylene chloride, 
n-propyl bromide, ethyl bromide, and ethyl alco- 

curs at a pressure of about 2,000 kg/cm 2 for 
water, at 12,000 for gallium, and at 25,000 
for bismuth. 

The phase changes of these three sub- 
stances afford a particular example of poly- 
morphism. The phase diagram of bismuth 
is shown in Fig. 9. Under pressure, poly- 
morphism is a very common phenomenon; 
the number of instances increases with in- 
crease in the experimental pressure range 
and with increasing sensitiveness in the 
methods for detecting small discontinuities 
of volume. In the range from room tempera- 
ture to 200° C and up to pressures of 50,000 
kg/cm 2 , roughly one-third of the substances 
examined have proved to be polymorphic. 
In the much greater range of conditions en- 
countered in the crust of the earth, the pre- 
sumption seems to be that no substance 
exists in the lattice with which we are fa- 
miliar under laboratory conditions, unless 
perhaps the lattice is of a particularly sim- 
ple type. The importance of such a con- 
clusion for geophysics is obvious. 

The thermodynamics of a polymorphic 
phase change is the same as the thermody- 
namics of melting, but beyond that there is 
little resemblance between the two phe- 
nomena; there is no such thing as "a" poly- 



vol. 38, no. 5 

morphic transition curve as there is "a" 
melting curve. There are only three falling 
melting curves, and these disappear at 
higher pressures; there are many falling 
transition curves, and they become increas- 
ingly numerous at higher pressures. Be- 
tween 12,000 and 50,000, 41 percent of the 
new transition curves are of the falling type. 
Transition curves may have horizontal or 
vertical tangents; melting curves have 
neither. Transition curves may have up- 
ward or downward curvature; melting 
curves are always concave downward. The 
difference of volume of two polymorphic 
phases may increase or decrease in the direc- 
tion of increasing temperature along the 
transition line ; the difference of volume be- 
tween liquid and solid always decreases. 
The compressibility of the high pressure 
phase may be greater or less than that of 
the low pressure phase; the compressibility 
of the liquid is always greater than that of 
the solid. Substances are capable of existing 
in a number of polymorphic forms, and the 
complete mapping of the transition tem- 
peratures and pressures for all the forms 
may result in phase diagrams of great com- 

plication. Thus bismuth has six different 
phases; water, which has some striking 
analogies to bismuth, has seven phases. The 
most complicated phase diagram investi- 
gated to date is that of camphor, which has 
eleven phases. 

There are only two generalizations with 
regard to polymorphic transitions that 
stand to date. The first is that critical points 
and continuous transitions between differ- 
ent polymorphic forms do not occur. If 
there were such points this would demand a 
continuous transition from one type of lat- 
tice to another, and this seems highly im- 
probable, although perhaps not logically 
inconceivable. The second generalization is 
that transitions in the simple CsCl type of 
lattice in the direction of smaller volumes 
are not produced by pressure; this lattice 
seems to be of such a simplicity that it is not 
likely to be disturbed. This second generali- 
zation naturally rests on a much smaller 
number of examples than the first, and is 
correspondingly less secure. 

We have so far been discussing transi- 
tions which are thermodynamically re- 
versible; when pressure is released the origi- 



10,000 20,000 30,000 


Fig. 9. — The phase diagram of bismuth. The arrows on the transition line III— IV indicate the 
pressure limits within which the transition runs with increasing or decreasing pressure. 

May 15, 1948 bridgman: the field of high pressure physics 


nal form is resumed. In addition to these 
reversible transitions, the existence of irre- 
versible transitions is to be recognized, that 
is, of changes effected by pressure which re- 
main permanent when they have once taken 
place. Two well-marked examples of this 
have been found. The first was phosphorus. 
If ordinary yellow phosphorus is exposed to 
pressures above 12,000 kg/cm 2 at tempera- 
tures above 200° C, it is permanently 
changed into a black solid much like graph- 
ite in appearance and like it a conductor of 
electricity, although yellow phosphorus is a 
good insulator. This remained the only ex- 
ample for many years. Recently I have 
found that ordinary liquid CS 2 may simi- 
larly be changed permanently into a black 
solid at temperatures in the neighborhood 
of 200° C and by pressures of the order of 
40,000 kg/cm 2 . This black substance is defi- 
nitely not a mixture of sulphur and carbon, 
which one might at first expect, but is ap- 
parently a unitary substance, truly a black 
solid form of carbon bisulphide. It has been 
suggested that the structure may be that of 
a single giant molecule like the known struc- 
ture of Si0 2 , which from the atomic point of 
view is very similar. It is fascinating to 
speculate that there may be many other 
common substances which may be pushed 
by sufficiently high pressures over a po- 
tential hill of some kind permanently into 
some hitherto unknown form. Until we 
have theoretical understanding of these two 
known permanent transitions, we can not 
help attaching a certain reasonableness to 
the assumption of the possible existence of 
other such substances. In fact, there is 
experimental evidence that many other 
such transformations may be effected. In 
experiments in which I combined high 
shearing stresses with high hydrostatic 
pressure I have observed some cases of ir- 
reversible transitions to forms already 
known, and have also observed a large 
number of color changes, which are the indi- 
cation of some sort of permanent change. It 
was not possible to establish whether new 
substances were formed under these con- 
ditions because the quantities of material 
involved were too small to permit satisfac- 
tory analysis. 

We pass now to other sorts of pressure 

effects. Perhaps the simplest of these to 
measure are the effects of pressure on elec- 
trical resistance. Measurements have been 
made at room temperature or higher up to 
30,000 kg/cm 2 and at the temperature of 
liquid air to 7,000. At low temperatures 
there is a natural limit to the pressure range 
imposed by the freezing of the medium 
transmitting pressure, which in this case 
was gaseous nitrogen. Fig. 10 shows the ef- 
fect of pressure on the alkali metals at room 
temperature up to 30,000 kg/cm 2 . 

In the first place, there is a specific effect 
of pressure on resistance; the pressure co- 
efficient of resistance is in general of the 
order of magnitude of ten times greater than 
the volume compressibility. This would in- 
volve as a corollary that the effect of pres- 
sure on the resistance of a highly compres- 
sible metal is greater than on a metal of low 
compressibility. This is indeed true in 
general, but exceptions are common. The 
resistance of perhaps three-quarters of the 
metals decreases with increasing pressure; 
as is to be expected, the rate of decrease 
itself decreases with increasing pressure, 
that is, the curve of resistance against pres- 
sure is convex toward the pressure axis. On 
the other hand, there are several metals, of 
which lithium, strontium, and bismuth are 
examples, whose resistance increases under 
pressure. For these metals, surprisingly, 
there is a law of increasing returns, that is, 
the rate of increase of resistance itself in- 
creases with increasing pressure. This means 
that for these metals also the curve of re- 
sistance against pressure is concave up- 
ward. Finally, there are a few metals which 
combine both types of behavior, that is, 
the resistance initially decreases, then 
passes through a minimum, and then turns 
upward. Examples are caesium, rubidium, 
potassium, and barium. It would appear, 
therefore, that the upward curvature is 
common to all metals, and that all resist- 
ance curves may be regarded as pieces of 
one single curve, the only difference for 
different metals being that what might be 
called the intrinsic zero of pressure is dif- 
ferently situated with respect to atmos- 
pheric pressure for different metals. 

Considerable success has been achieved 
by theoretical physicists in explaining theo- 



retically the effect of pressure on resistance. 
As might be expected when effects depart- 
ing so largely from linearity are concerned, 
we can recognize the presence of at least two 
different mechanisms working in opposite 
directions. There is in the first place an 
effect of pressure on what is the analogue of 
the electron free path in the old electron gas 
theory of metallic conduction. This is con- 
nected with the change of dimensions, and 
in general works in the direction of an in- 
crease of free path, that is, a decrease of 
resistance, with increasing pressure. In the 
second place, there is a rearrangement of 
the energy levels, and this may, when the 
energy bands are nearly completely occu- 
pied, work in the direction of a change in 
the effective number of free electrons. De- 
eonding on the details of the atomic rela- 
tions, this effect may be either an increase 
or a decrease. The calculations have been 
carried through approximately in a few 

simple cases. It turns out that the increase 
of resistance of lithium with pressure is due 
to the preponderating effect of a decrease in 
the effective number of free electrons. 

The effect of pressure on the electrical re- 
sistance of single crystals is sometimes very 
complicated. If the crystal system is cubic, 
the material behaves electrically like an 
isotropic body, but if the system has lower 
symmetry, there may be differences in dif- 
ferent directions. In antimony, for example, 
the sign of the pressure effect is different in 
different directions. There are directions in 
the crystal for which the resistance passes 
through a maximum with increasing pres- 
sure, whereas for other directions the re- 
sistance decreases with normal curvature. 

The resistance of some semi-conductors 
may be decreased by such large amounts 
that they approach the absolute resistances 
characteristic of the metals. An early in- 
vestigation in this field was made on 










10,000 \5,000 


Fig. 10. — The relative resistances of the alkali metals up to 30,000 kg/cm 2 . The break in the curve 
for caesium is due to a polymorphic transition. Potassium has a very flat minimum near 23,000. 

May 15, 1948 



selenium and silver sulphide by Mont6n in 
Uppsala. At higher pressures, tellurium 
approaches the properties of a metallic con- 
ductor under 30,000 kg/cm 2 . Not only does 
the absolute value of the resistance drop to 
a characteristically low figure, but the tem- 
perature coefficient, which initially is nega- 
tive, reverses sign under high pressure and 
becomes positive like that of the metals 
proper. Theory is as yet hardly in a position 
to explain these complicated effects, either 
in single crystals or in semi-conductors. 

Closely related to the electrical conduc- 
tivity of metals is their thermal conductiv- 
ity; the relationship is expressed by the 
approximate equality of the Wiedemann- 
Franz ratio of electrical to thermal con- 
ductivity for all metals. Under pressure, 
thermal conductivity changes as well as 
electrical conductivity. It is much more dif- 
ficult to measure than electrical conductiv- 
ity, and satisfactory measurements have 
been made only for a few metals and those 
up to only 12,000 kg/cm 2 . It appears that 
for these metals the Wiedemann-Franz ratio 
is approximately independent of pressure. 

The effect of pressure on the thermal con- 
ductivity of liquids is much larger than on 
that of metals, and is much easier to meas- 
ure. In general, the thermal conductivity 
increases for ordinary liquids under a pres- 
sure of 12,000 kg/cm 2 by a factor varying 
between 2 and 3. The effect on water is 
smaller; at 12,000 the increase for it is only 
50 percent. There is a close connection be- 
tween the effect of pressure on thermal con- 
ductivity of normal liquids and the effect of 
pressure on the velocity of sound in the 
liquid. That is, thermal conductivity in a 
liquid is primarily a mechanical affair; heat 
is transferred by microscopic mechanical 
waves travelling with the velocity deter- 
mined in the conventional Way by the com- 
pressibility. The small effect in water is 
associated with the small change in the 
compressibility of water brought about by 

Another property of metals obviously re- 
lated to electrical and thermal conductivity 
is the thermo-electric characteristics. These 
properties are also affected by pressure. In 
general, a metal under pressure behaves 
thermgelectrically differently from the same 
metal not under pressure, so that a thermo- 




5000 I OCjOO 

Pressure, Kg/c/h* 

Fig. 11. — The effect of pressure on the 
viscosity of i-butyl alcohol. 

couple may be made with one branch of any 
ordinary metal and the other branch of the 
same metal exposed to hydrostatic pressure. 
Under a pressure of 12,000 kg/cm 2 the 
thermoelectric power of such couples may 
be as large as that of ordinary couples com- 
posed of two entirely different metals. A 
number of such "pressure couples" have 
been investigated. The effects are compli- 
cated; there is not any universal rule with 
regard to the sign of the effect. There may 
be reversals of sign and large departures 
from linearity. No satisfactory theory of 
these effects has been formulated. At pres- 
ent one can only draw the conclusion that 
the interplay of several different mecha- 
nisms must be involved. 

The largest of all the pressure effects 
studied to date is on the viscosity of liquids. 
In general, viscosity increases under pres- 
sure at a rate increasing rapidly with in- 
creasing pressure. The curve of viscosity 
against pressure usually rises exponentially 
with pressure and sometimes more rapidly 
than exponentially. In Fig. 11 is shown 



the viscosity of i-butyl alcohol at 30° and 
75° at pressures up to 12,000 kg/cm 2 . The 
total rise may be by as much as a factor of 
10 7 for a pressure increase of 10,000 kg/cm 2 
(for eugenol). The rate of rise is definitely 
linked with the complication of the mole- 
cule, and is greater the more complicated 
the molecule. For the comparatively simple 
liquid water the rise of viscosity under 

10,000 kg/cm 2 is by a factor of only 2 or 3 
and for monatomic mercury by only 30 per- 
cent. For methyl alcohol the increase is by 
a factor of 10, for propyl alcohol by a factor 
of 100, and for amyl alcohol by a factor of 
1,000. In the last few years the theoretical 
physicists have had considerable success in 
accounting for the effect of pressure on the 
viscosity of liquids. 


Readers of Professor Bridgman's Nobel 
Lecture may be interested also in his re- 
marks on the subject "Science and Free- 
dom" that he made on January 11, 1947, at 
a dinner given in his honor by the Dean of 
the Faculty of Arts and Sciences of Harvard 
University at the Harvard Club of Boston. 
The address is here reprinted from the Isis 
37: 128-131. 1947, by the kind permission 
of Dr. George Sarton, editor of that journal. 
In introducing Professor Bridgman's re- 
marks in the Isis, Dr. Sarton wrote: "They 

are of great interest to historians of science 
because they reveal the afterthoughts which 
do not appear in [his] scientific papers, nor 
even in the philosophical ones, yet which 
are essential for the understanding of a 
man's personality. The document ... is of 
great value for the 'new humanists/ who 
wish to understand not only the technical 
aspects, but also the human factors, with- 
out which there would be no science, or 
without which science would lose its mean- 
ing and its grace." 

By Percy W. Bridgman 

This will not attempt to reproduce exactly 
what I said at the dinner on January 11, 1947, 
but I shall avail myself of the suggestion of 
Dr. Sarton to make a partly imaginary speech, 
composed of parts of what I actually said, and 
of what, in the light of afterthought, I wish I 
had said. I shall not attempt to reproduce a 
number of the more or less personal and in- 
formal details, but shall confine myself to mat- 
ters of more general interest. 

Of all the conditions of my work which in 
retrospect appears most important, and of 
which at the time also I was keenly conscious, 
freedom of investigation is outstanding. There 
has never been any suggestion from any out- 
side source as to the nature of my investiga- 
tions. Even in the early days, when I sought 
and obtained the maximum relief from teach- 
ing and administrative duties for the ostensible 
purpose of more complete devotion to my re- 
search, no attempt was made by the University 

authorities to impose as a condition that I 
continue to devote myself to high pressure in- 
vestigation or even to investigation itself. The 
apparent attitude of the authorities was that 
if you are going to gamble that you have found 
a good man, a gamble without strings attached 
is the most likely to succeed. Any consistency 
which my experimental program may have 
shown has been a consistency imposed entirely 
from within; this I believe to be the proper 
source of consistency. In spite of the fact that 
I have in the main followed one guiding experi- 
mental idea, I have nevertheless at all times 
felt free to pursue other lines of interest, wheth- 
er experiment, or theory, or fundamental 

Another outstanding characteristic of my 
work has been the smallness of its scale. Not 
only is the apparatus itself small, in fact be- 
coming smaller the higher the pressure, because 
of inherent physical limitations on strength, 

May 15, 1948 

bridgman: science and freedom 


but I have never had more than two or three 
students at a time or a couple of assistants. 
The result has been that I have been able at all 
times to maintain the closest contact with the 
details of the work, and also have been able to 
conserve the requisite amount of leisure. Both 
of these features have been of the highest im- 
portance. In advancing into new territory, as 
in this high pressure work, the necessity is con- 
tinuous for the development of new methods 
and new ideas. For me, at least, new ideas 
germinate only in an atmosphere of leisure. I 
have to immerse myself in a problem and then 
let it gestate in my brain, without the distrac- 
tion of other interests, if I am to expect the 
solution to come sauntering into my mind 
when I wake up two or three mornings later. 
In this process manual cooperation plays a 
great part. Adjoining my laboratory is my 
machine shop; in fact, it is an integral part of 
the laboratory to which I can repair and stimu- 
late inspiration by working out half-formed 
ideas with my own hands. Not only do I have 
enough leisure so that I can work in the shop 
with my own hands on occasion, but I am also 
able to carry through my own experiment, in- 
cluding making all the readings, myself. I find 
this necessary if I am to have confidence in the 
results of some method not hitherto tried. 
There are too many pitfalls of unanticipated 
sources of error, which often require ingenuity 
for their elimination, and which may take 
much time to discover if one is only watching 
from the side lines. I have been able to make 
it an invariable practice to stay with each new 
method long enough to get material for a com- 
plete paper, before turning the method over to 
an assistant for more or less routine application 
to a large number of substances. Not only this, 
but even when an assistant makes the experi- 
ment and the readings, I have always made the 
computations and written the paper myself. 
This gives me a confidence in the results not 
possible when working on a larger scale. An- 
other great advantage of working on a small 
scale is that one gives no hostages to one's 
own past. If I wake up in the morning with a 
new idea, the utilization of which involves 
scrapping elaborate preparations already made, 
I am free to scrap what I have done and start 
off on the new and better line. This would not 
be possible without crippling loss of morale if 
one were working on a large scale with a com- 
plex organization under one. 

Another characteristic of the field in which I 
have been working is that it is not a particularly 
popular field, so that there have been com- 
paratively few workers in it and corresponding- 
ly little competition. This has both advantages 
and disadvantages. It is an advantage that one 
can do his work with no sense of hurry, so that 
there is little temptation to make premature 
announcements, and should questions arise 
one can take the time to repeat the experiment 
or make other modifications that will clear up 
the matter. Also, the order in which the prob- 
lems are attacked can be the order of greatest 
scientific economy, rather than the order of a 
competitive politics. On the other hand, the 
principal disadvantage, obvious enough to 
everyone, is that the investigator loses the 
stimulation of conversation with his colleagues 
on mutual problems. Just how important this 
stimulus is will depend in considerable measure 
on the individual investigator; some may find it 
well nigh indispensable, whereas others may be 
much less dependent on it. I myself have been 
able to get along in considerable measure with- 
out it. Even at scientific meetings, which every 
physicist seems to have to attend at intervals 
for rehabilitation of his inner man, the stimulus 
which I have received has not been detailed 
and specific, but rather general, in suggestions 
of trends and areas of coming interest. It has, 
I think, been a happy circumstance that my 
field, although obviously narrow in the sense 
that pressure is a highly specialized physical 
parameter, nevertheless from another point of 
view has been exceedingly broad. For the gen- 
eral problem has been no less than to deter- 
mine the effect of pressure on all physical prop- 
erties, and it therefore covers the entire reach 
of physical phenomena with the exception of 
such things as vacuum tube phenomena. 

Mention of the stimulus of conversation with 
one's colleagues naturally prompts one to con- 
sider the increasing trend during the last few 
years to large-scale cooperative enterprises 
among physicists. The reasons for this are 
obvious in the enormously increasing size and 
expense of the apparatus necessary for modern 
physical research, such as the cyclotrons and 
the piles of nuclear physics. Although we may 
recognize that such instruments are necessary, 
we may nevertheless deplore some of the conse- 
quences. Up to now ideas have been in such a 
rapid state of flux that the instrument itself 
has been continually evolving, with the result 



that most physicists in this field have been 
spending an increasingly large fraction of their 
tiir.e on the purely engineering job of the design 
and construction of new and better instruments 
and correspondingly an increasingly small pro- 
portion of time on the calculation of results 
and rumination on their significance. The 
competition in this field is intense; rivalry 
between different groups at different universi- 
ties can offer little opportunity for leisure or the 
scholarly digestion of results before publication. 
Within the last year there has been one glaring 
example of hasty publication of a spectacular 
result of such presumptive importance as to 
start a rush of other investigators into the field, 
only later to be withdrawn as erroneous be- 
cause of inadequate consideration of factors 
which obviously were crying for evaluation in 
the beginning. Not only is there haste because 
of competition, but there is haste because of 
financial considerations. The apparatus is so 
expensive that consideration of the overhead 
demands that the apparatus be kept in opera- 
tion for twenty-four hours a day, and this is 
not conducive to a feeling of leisure. Each of 
the teams which is the slave of one of these in- 
struments has to be driven by some one at the 
head who has the ideas. There is danger here 
that all the rest of the team will pick the brains 
of one man, with an ultimate decrease in the 
number of physicists in the community capable 
of independent and critical thought. Still worse, 
the physicist who should be directing his team 
by his creative ideas is likely to be so swamped 
by the administrative details of the large en- 
terprise under him that he is overwhelmed and 
his purely scientific activity destroyed. This is 
well known to have happened, at least tem- 
porarily, in the case of one of the new mammoth 
calculating machines, machines whose osten- 
sible purpose is to free the scientist from drud- 
gery and make possible the creative use of his 
time. Doubtless some physicists have the nat- 
ural knack of being able to work together 
harmoniously and perhaps even efficiently in 
teams, and perhaps others can acquire it, but 
I believe there are many who are permanent^ 
unfitted for effective cooperation in this way, 
and it will be a major loss if they are not able 
to find a niche in which thejr may function. 

During the war practically all the physicists 
in this country were diverted to war work of 
one sort or another, and a large part of them 

were engaged in large-scale enterprises which 
involved team work developed to its maximum 
efficiency, with the consequent and necessary 
submergence of the individual. The older men, 
who had previously worked on their own prob- 
lems in their own laboratories, put up with this 
as a patriotic necessity, to be tolerated only 
while they must, and to be escaped from as soon 
as decent. But the younger men, who had been 
drawn into the maelstrom before starting work 
for their Ph.D. degree, had never experienced 
independent work and did not know what it is 
like. Some of these younger men will continue 
in government work; others who return to 
academic circles will there join in the teams 
serving the mammoth instruments. The result 
is that a generation of physicists is growing up 
who have never exercised any particular degree 
of individual initiative, who have had no op- 
portune to experience its satisfactions or its 
possibilities, and who regard cooperative work 
in large teams as the normal thing. It is a 
natural corollaiy for them to feel that the ob- 
jectives of these large teams must be something 
of large social significance. The temper of the 
rising generation is recognizably different from 
that of the older. I may mention one example 
with which I have had personal acquaintance. 
The Association of Cambridge Scientists was 
one of many similar associations formed soon 
after the dropping of the atomic bomb on Hiro- 
shima to consider all the implications of the 
situation thus created. In the early daj^s of the 
Association the May- Johnson bill was a matter 
of much concern. With regard to this there was 
in the ranks of the Association a cleavage of 
attitude almost exclusively along lines of age. 
The older men were troubled and concerned 
by the threats to scientific freedom contained 
in the bill, whereas the younger men were not 
at all concerned about this, but took the oppo- 
site view that it was on the whole a rather base 
and self-indulgent thing for the individual not 
to be willing to sacrifice his scientific freedom 
on the altar of the good of society. The young 
men, never having experienced scientific free- 
dom, did and could not see that the question of 
self-indulgence does not enter at all into the 
situation, but the existence of science itself, 
which I think all conceded to be a social good, 
is impossible without scientific freedom. 

The increasing amount of administrative 
work falling on some of the scientists compos- 

May 15, 1948 



ing the large teams has a parallel in the ever 
increasing amount of routine administrative 
work expected from the members of a Univer- 
sity faculty. As I look back on the forty years 
of my work there can be no question but that 
the fraction of the time of the average faculty 
member spent in routine of administrative or 
other sorts has increased unconscionably. 
There seems to be a natural law operating here ; 
the larger an institution becomes, the more 
cumbersome and less efficient it becomes. One 
might think that when the institution becomes 
ten times as large it would have ten times as 
much business and ten times as many people 
to do it, so that each individual would need to 
give only the same amount of time. But it does 
not work this way; it is more like the develop- 
ment of a telephone exchange, which when it 
increases by a factor of n has to provide for the 
handling of factorial n as many combinations. 
Each new functionary in a University has to 
justify himself, which he does by exacting at- 
tention from every member of the faculty. 
When, for example, the office of archivist is 
created, every member of the faculty is asked to 
provide material for the archives. Ten times as 
many officials tend to demand ten times as 
much attention from each member of the 
faculty. In a recent number of the Scientific 
Monthly there was an amusing and satirical 
article depicting the eventual extinction of the 

human race by suffocation in its own intel- 
lectual effluvia. It is well known that every 
large library tends to increase in size geometri- 
cally with the simple arithmetical passage of 
time. Why this should have to be true does not 
seem to have been explained, but it is an un- 
doubted description of the observed behavior 
of the human animal. Unless some way can be 
discovered of breaking the cycle, the logical 
final result is catastrophic. In the past the cycle 
has been broken by wars and the collapse of 
civilizations. In the hoped-for brighter future 
in which war has been eliminated, some specific 
means will have to be devised to cope with the 
situation. The satirical article portrayed the 
consequences of the inability of the human race 
to devise an adequate means. It is no less a 
problem to devise a means by which the time 
of the members of university faculties may be 
saved for creative effort. Otherwise creative 
science will be driven to other asylums, if in- 
deed it is not destroyed. 

As I look to the future I am therefore trou- 
bled by two misgivings : that there will be less 
and less place for the small individual experi- 
menter, and that the time of all of us will be 
increasingly commandeered by administrative 
mechanical details. In view of these misgivings 
I cannot help wondering as I look back on the 
past whether, if I were to start over again now, 
I would be able to do again what I have done. 

CHEMISTRY.— A method for the determination of certain metals present in minor 
concentration in various substances. 1 N. Howell Furman, C. E. Bricker, 
and Bruce McDtjffie, Princeton University. (Communicated by James 
I. Hoffman.) 

This paper presents an account of a 
systematic scheme for the determination 
of certain metals in a variety of substances. 
The scheme at first was directed toward the 
estimation of minor amounts of a few other 
metals that were found to occur in uranium 

1 Lecture delivered March 13, 1947, on the oc- 
casion of the award of the Hillebrand Prize in 
Chemistry by the Chemical Society of Washing- 
ton to Dr. James I. Hoffman in recognition of his 
contributions on the extraction of alumina from 
clay and the purification of uranium for atomic 
energy. Received December 31, 1947. 

Parts of the work briefly reported on herein 
were done under O.S.R.D. contracts NDRC-121 
and OEM-sr-649 and Manhattan Project Con- 
tract W-7405-eng.-81, and this support is grate- 
fully acknowledged. 

or its compounds and salts. The method 
was gradually modified and extended in 
scope from 1942 to 1945 with the aid of 
G. P. Haight, Jr., J. A. Nyitrai, and others. 
The proved applications of the process, in 
addition to those already mentioned, in- 
clude the estimation of minor amounts of 
various metals in the salts, alloys, or nearly 
pure metallic specimens of Ba, Be, Ca, Mg, 
and Na. From the experience thus far ob- 
tained it appears probable that the method 
could be applied to the testing of many 
other pure metals, their alloys, or com- 
pounds. The method is also applicable to 
the collection of the traces of metals that 
may occur in organic compounds or in 



materials of biological and pharmaceutical 

The original problem had to do with the 
detection or the determination of certain 
metals of the hydrogen sulphide group, 
notably copper, cadmium, lead, and molyb- 
denum, or the detection of copper, cadmium, 
iron, etc., in lead and its salts, or in bismuth 
and its compounds. It was therefore natural 
to explore the possibilities of mercury cath- 
ode electrolysis as a preliminary to polarog- 
raphy and colorimetry. The pioneer work 
of E. F. Smith on mercury cathode separa- 
tions was at once utilized (1942) in connec- 
tion with the estimation of minor amounts 
of cadmium in lead or bismuth or in lead 


The majority of the procedures that were 
later developed in detail centered around 
the use of a small mercury cathode as a 
collector for minor amounts of metals which 
were electrolyzed into the cathode under 
the conditions chosen. The mercury cath- 
ode was then drained into a weighed silica 
combustion boat without interrupting the 
current, and after removal of the mercury 
by distillation in a current of nitrogen the 
residues were weighed and dissolved and 
the conditions were adjusted for the ap- 
plication of polarography. After the polaro- 
graphic estimations the residual solution 
was further utilized for certain spectro- 
photometric determinations. The solution, 
from which the electrodepo sited metals 
had been removed by electrolysis, was ex- 
amined systematically by extractions and 
colorimetric procedures for minor amounts 
of other metals, e.g. Mo, Fe, Ti, and V (1). 

This method of concentration has very 
decided advantages prior to polarography 
because the supporting electrolyte can be 
controlled very simply for the estimation of 
minor amounts of metals recovered from a 
great variety of samples. The method has 
been found useful for the concentration of 
materials prior to spectrophotometric or 
spectrograph^ estimations. For routine ap- 
plication to a given material the weight of 
the residue from the mercury distillation 
gives an idea of the gross contamination 

of the material by the electrolyzable metals. 
Chance contamination of some part of a 
sample can be detected by virtue of the 
fact that a very large sample may be 

The general facts of electrolysis into a 
mercury cathode have been summarized 
as follows (2): The following metals singly 
or in association are quantitatively de- 
posited in a mercury cathode: Cu, Ag, Au, 
Zn, Cd, Hg, Ga, In, Tl, Ge, Sn, Bi, Po, Cr, 
Mo, Re, Fe, Co, Ni, Rh, Pd, Ir, Pt. Further 
Pb, As, Se, Te, and Os are quantitatively 
separated from the solution, but are not 
necessarily deposited in the mercury. Mn, 
Ru, and Sb are partially separated from the 

The foregoing general facts had not been 
studied exhaustively in the minor concen- 
tration ranges in which we were interested, 
nor had the recovery of various metals by 
electrolysis into mercury and distillation of 
the latter been studied, although the re- 
covery of gold after amalgamation is a 
well-known process. 

Oxidation-reduction buffers. — In the elec- 
trolysis of specimens of which the major 
constituents may form oxidation-reduction 
systems such as the U IV -U m , Ti IV -Ti m , 
V m — V 11 , etc., the list of metals that are 
completely deposited in the mercury is cur- 
tailed because the cathode potential is not 
depressed far enough to reach the level 
necessary to deposit some of the metals. For 
example, from rather concentrated solu- 
tions of uranium sulphate no Mo, Cr, and 
Mn are deposited under conventional con- 
ditions of electrolysis. A further point that 
we have noted is that the level which the 
cathode reaches is affected by the deposition 
of traces of platinum, iron, copper, etc., on 
the mercury. These and other metals tend 
to lower the overvoltage of hydrogen depo- 
sition and to make the removal of certain 
metals less complete. Iron under some con- 
ditions is completely removed from a solu- 
tion containing uranium but from concen- 
trated solutions the removal is uncertain 
and it is necessary to test for iron both in 
the mercury deposit and in the aqueous 
solution after electrolysis. 

The details of electrolysis, distillation, 
and colorimetry can best be described in the 

May 15, 1948 



following outline of the experimental de- 
velopment : 


The scale of operation was set up prima- 
rily for the analysis of micro-amounts of 
metals. The apparatus was so chosen that 
distillation residues in the range from 10 
micrograms up to 5 mg of material could be 
handled. The sample size is adjusted so that 
the material electrolyzed and left after dis- 
tillation will be in this range. This scale of 
operation is well suited to polarography. In 
general residues between a few tenths of a 
mg and 1.5 mg were handled so that the 
process is a true microprocedure from this 

Apparatus and materials. — Silica dishes 
of 250 ml capacity or smaller were used in 
the preparation of solutions except for 
fluorides which were treated in platinum 

The reagents used to dissolve the samples 
were at first redistilled from a silica still for 
acids or from Pyrex for distilled water. Re- 
distilled mercury was used. Later it was 
found feasible to use analytical reagent 
acids and to run frequent reagent blanks 
through the whole process. 

Electrolysis cells of special design as 
shown in Fig. 1 were used. It is important to 
have the wire leading to the cathode termi- 
nate as near the stop-cock as possible. 
Anodes of platinum alloyed with 10 percent 
of iridium are satisfactory. The mercury 
cathode is drained into a silica combustion 
boat 14 by 75 mm and 10 mm deep, with 
handle. Silica beakers of 150 ml capacity 
are used for dissolving the deposit. 

A 12-inch split tube furnace of conven- 
tional design is used for the distillation. 
The nitrogen that is used to sweep out the 
mercury is not freed from air. 

A recording polarograph, preferably of 
the Heyrovsky type is desirable. The instru- 
ment that was used in the development of 
the method has been described in the litera- 
ture (5). 


The initial experiments were made by 
electrolyzing exhaustively a solution con- 
taining a large amount of the substance to 

be tested. The first mercury cathode was 
drained out and washed out with a little 
mercury. Then with a fresh mercury cath- 
ode a known addition of any metal to be 
tested was made, and the electrolysis was 
repeated and the mercury was drained out, 
distilled, and the residue tested. Since cad- 
mium did not occur in the reagents, as 
proved by frequent blank tests, the extent 
of recovery of varying amounts of cadmium 
may be used as an illustration of the effec- 
tiveness of the method, Table 1. 

Table 1. — Typical Recoveries by the Procedure 













0.095 -0.100 



0.45 -0.50 




















* Where a range is given several individual determinations 
were made. 

A number of similar experiments were 
made with small amounts of various sub- 


^ Platm urn 

Fig. 1. — Cell for electrolysis. The cell is pro- 
vided with a lip not shown. The anode is inserted 
over the lip and a cover glass is used. 



VOL. 38, NO. 5 

stances added to uranyl solutions that had 
been purified by electrolytic or other meth- 
ods. The results are corrected for blank 
values, and typical data are given in 
Table 2. 

Table 2.- 

-Recoveries of Various Elements 
by the Procedure 




Method of 
























about 100* 




















None deposited, owing to oxidation-reduc- 


tion buffering action. 



Doubtful. Probably none deposited. 




Colorimetric. In general 
Mn is electrolyzed into 


* Several individual determinations were made. 

A superficial investigation was made of 
the behavior of Bi, In, Tl, Ir, Ru, Ga, Au, 
Pd, Re, Se, and Te. The first three are 
quantitatively taken into the mercury. It 
was not definitely established that the re- 
maining substances can be deposited in the 
mercury, although Ga, Au, Pd, and Re gave 
distillation residues that indicated sub- 
stantial recoveries. 

Blanks on reagents. — In routine opera- 
tion, blanks are run each time a new bottle 
or other supply of any reagent is introduced. 
A number of sealed containers of reagents 
of the same lot are set aside for the work 
that is planned. A typical set of reagents for 
handling many types of samples is as fol- 
lows : 

Mercury, 2.5 ml. 

Water, distilled, 160-165 ml. 

Nitric acid, cone. 6 ml. 

Maximum supressor, 0.1 ml. 

Sulphuric acid, cone. 5.35 ml. 

Hydrochloric acid, cone. 2.1 ml. 

Potassium chloride, 25 mg. 

Pyridine 0.15 ml. 

Successive blanks obtained over a period 
of several weeks had the following range, in 
milligrams, Cu ; 0.011 to 0.026; Fe, 0.023 to 

0.039; Ni, 0.001 to 0.002; and Pb, 0.002 to 

Cadmium, cobalt, and zinc were not en- 
countered in these reagents. Fluctuations in 
copper blanks were in one instance traced 
to a copper pin in an atomizer bulb that was 
used with a stock bottle of nitric acid. A 
series of erratic results for cadmium was due 
to cadmium contained in methanol that 
was used to wash metallic specimens that 
were greasy. The cadmium content of the 
methanol was 2 parts per million. It proved 
to be unwise to use the polarographic cells 
and the silica beakers for experiments with 
major amounts of any of the metals that 
were being estimated. The iron blank in- 
cludes additional iron introduced by a suc- 
cession of reagents that were used in con- 
nection with the recovery of iron from the 
solution that had been electrolyzed. 

Calibrations. — Although it was possible 
to control the composition of the solution 
that was subjected to polarographic analy- 
sis rather rigorously, it was considered de- 
sirable to calibrate with mixtures of ele- 
ments in the normal ranges that were 
encountered in applying the method. Cer- 
tain calibrations were made with materials 
put through the complete process and with 
proper blank deductions. Other calibrations 
were made with mixtures of elements taken 
from standard solutions of their chlorides or 
nitrates. The solutions were handled in 
exactly the same fashion as the residues 
from the distillation of mercury. Perhaps 
the best evidence that the calibrations were 
consistent is to be found in the data of 
Tables 1 and 2. Some of the determinations 
there reported were on single substances 
and others were for mixtures of Cd, Cu, Co, 
Ni, Zn, and Fe in varying ratios. 

Reproducibility of the procedure. — A few 
typical results on the recovery of minor 
amounts of elements from solutions are 
shown in Table 3. 

Approximately the same degree of re- 
producibility was obtained in check deter- 
minations on numerous samples of various 
types. A typical example of the polaro- 
graphic findings is shown in Fig. 2, Avhich 
presents the polarograms taken from solu- 
tions of two equal samples of a slightly 
alloyed light metal. 

May 15, 1948 furman et al.: determination of certain metals 


^ FlG .V 2 -~;?-! 1 - plica 5 e analyses Curves 1 and 2 are taken before adding pyridine; curves 3 and 4 
metal n Py ndme - The z * n c wave in curve 4 is used for the qualitative estimation of this 

I til 



A sample is taken of suitable size to give 
a residue in the range of tenths of a milli- 
gram to 2 mg after electrolysis and distilla- 
tion. The samples may range from a few 
tenths of a gram up to 20 g or more with 
the equipment specified. If the material will 
dissolve in sulphuric acid, an amount calcu- 
lated to leave 0.5 ml of the free concentrated 
acid in excess is used. If nitric acid is 
needed, a measured minimal amount is 
used. With samples containing much cal- 
cium, it is desirable to use hydrochloric acid 
followed by sulphuric acid and evaporation 
and fuming to convert to sulphates. Alter- 
natively for samples containing calcium 
perchloric acid may be used during the solu- 
tion and electrolysis. 

Uranium and its compounds are con- 
verted to sulphates, usually after prior 
treatment such as burning the metal to 
oxide and treatment with sulphuric acid. It 
is important to fume off the free sulphuric 
acid and then to add prior to the electroly- 
sis an amount of free sulphuric acid equiva- 
lent to 1 mole of H2SO4 per gram atom of 
uranium because there is an average con- 
sumption of 1 mole of acid during the con- 
version of one gram atom of U^ 1 to U m . 
Two moles of acid are used up in the reduc- 
tion from state VI to state IV and one mole 
of acid is regenerated in the reduction of 
U IV to U m . If the acidity is not regulated, 
the recovery of the desired metals may be 
incomplete. If too little acid is present, hy- 
drolysis may occur during the electrolysis. 

A reagent blank has to be carried through 
with any type of process. 

If there is an insoluble residue, it may be 

Table 3. — Duplicate Analyses of Solutions for Small 
Amounts of Metals, in Milligrams of Elements Found 







1, a. 














2, a. 














3, a. 













4, a. 














filtered and if silica is present, its weight is 
determined by hydrofluoric acid treatment 
and the residual oxides are dissolved and 
added to the filtrate. 

Electrolysis. — A 2.5 ml mercury cathode 
is placed in the cell, Fig. 1, and the solution 
is added and diluted to 100 ml. A platinum- 
iridium (10 percent) anode is inserted to a 
depth of 1.5 cm. A wire of B. and S. gauge 
15 (0.058 inch diameter) is satisfactory. 
The cell is covered with a watch glass. With 
an applied voltage of 10 volts the series re- 
sistance is regulated so that 0.8 ampere 
flows. This corresponds to an initial cathode 
current density of about 0i08 ampere per 
cm 2 . At the end of the electrolysis the 
current has usually dropped to about half 
the initial value. In order to remove the 
more active metals such as zinc it is desira- 
ble that the cathode potential shall go down 
well into the region of the lowest soluble 
oxidation-reduction system that may be 
present. In the case of uranium it is desira- 
ble to have as much trivalent uranium 
present as possible. Mechanical stirring 
causes much of the U m to be oxidized to 
U IV at the anode or at the exposed sur- 
face. Such stirring does not interfere with 
the complete removal of iron. Lead if pres- 
ent in minor amounts goes completely into 
the cathode when an oxidation-reduction 
buffer is present. Iron is completely electro- 
lyzed out of solutions that have no oxida- 
tion-reduction buffering action. With many 
types of samples the small mercury cathode 
is observed to rotate spontaneously at a 
slow rate that gives a certain desirable 
amount of agitation near the cathode. 

Distillation. — After the electrolysis, 
which is normally allowed to proceed from 
late afternoon to the next morning, or from 
early morning to late afternoon, the mer- 
cury cathode is drained into a weighed silica 
boat, without interrupting the current. The 
last drop of mercury is caught on a dry 
filter paper to free it from any trace of solu- 
tion, and then added to the boat. The 
aqueous solution is quickly washed into a 
beaker and the cell is inverted and tapped 
over a filter paper to recover the last glob- 
ules of mercury which are blotted dry and 
added to the boat. If there is as much as a 
few tenths of a milligram of other metals 

May 15, 1948 furman et al. : determination of certain metals 


present, the mercury has a visible surface 

The temperature of the boat is observed 
during the distillation with the aid of a 
thermometer inserted through the inlet- 
stopper of the tube furnace. A current of 
nitrogen, not freed from oxygen, is passed 
through the furnace during the distillation, 
and the mercury is condensed in a beaker of 
water into which the outlet tube dips. As 
many as three boats at a time are handled 
in a 12-inch furnace. After distillation the 
temperature is lowered and when the boat is 
nearly at room temperature a cylinder of 
clean paper is inserted in order to avoid 
contact of the boat with condensed mer- 
cury droplets. There is a characteristic 
change from a golden brown color to bluish 
with amalgams rich in cadmium and the 
temperature is lowered as soon as this 
change becomes evident.. If little or no zinc 
is present and the residue is largely iron it 
will be distributed in small specks on the 
bottom of the boat. Zinc and indium form 
amalgams which are decomposed only at a 
much higher temperature. The zinc amal- 
gam forms a continuous layer. The weight 
of the residue is of little significance when 
zinc is a major component of the residue. 
Indium was only encountered in synthetic 

In residues that contain primarily iron 
plus very minor amounts of copper, lead, 
etc.,. the weight of the residue less the Fe 2 3 , 
that is equivalent to the iron, agrees fairly 
closely with the sum of the metals other 
than iron in about 65-75 percent of the 
cases. The weight is therefore only a rough 
index of over-all content of electrolyzable 
metals. The average ratio of weight of de- 
posit to weight recovered polarographically 
and colorimetrically in the residue is about 
1.88 for a given class of material. The ratio 
Fe 2 3 /2Fe is 1.43. The discrepancy is in 
part due to the presence of minor amounts 
of mercury or platinum and to the fact that 
the elements other than iron probably are 
converted more or less completely to oxides. 

Solution of the deposit. Polarography. — 
The boat is placed handle end up in a silica 
beaker of 150 ml capacity, and the residue 
is treated with 2 ml of concentrated hydro- 
chloric acid and 1 ml of concentrated ni- 

tric acid. The acids are caused to come 
into contact with all the residue and 
the covered beaker is heated until solu- 
tion is complete. The boat is then washed 
with a minimum amount of water and re- 
moved. The solution is evaporated to dry- 
ness. Then 0.5 ml of concentrated hydro- 
chloric acid and an aliquot of a potassium 
chloride solution containing 25 mg of KC1 
are added and the evaporation is repeated. 
If mercury is known or suspected to be 
present, it is volatilized by heating on a hot 
plate at medium heat. The residue is finally 
evaporated after the addition of another 
0.5 ml of concentrated hydrochloric acid 
and a little water, at 100°C. The heating is 
stopped while the odor of HC1 is faint but 
present in the dry residue. This step is ve^ 
important because sufficient acid remains in 
the residue to prevent loss of iron or other 
metals. The residue is dissolved in 2.9 ml of 
water, and 0.1 ml of 0.1 percent methyl cel- 
lulose is added as a maximum supressor. 
The solution is made homogeneous and as 
much as possible is transferred to a polaro- 
graphic cell of the type devised in this labo- 
ratory (10). The solution is deaerated with 
nitrogen for 10 minutes, the nitrogen being 
passed through a purifying train and a wash 
bottle containing water. 

A polarogram is taken at the highest 
feasible sensitivity, recording that portion 
of the current voltage curve immediately 
following the spontaneous initial wave due 
to iron, etc. In this medium only the cad- 
mium and lead waves are useful. A second 
polarogram is taken after adding one drop 
of hydrochloric acid and three drops of 
pyridine and deaerating for a few minutes; 
the pH of the solution is about 5.2. Lingane 
and Kerlinger (4) proved that the precipi- 
tate of hydrated ferric oxide does not carry 
Cd, Cu, Co, Ni, or Zn. In this second 
polarogram the copper is determined from 
the Cu 1 to Cu° wave and a check is ob- 
tained as to the amount of cadmium. The 
nickel and cobalt waves are well separated, 
but the cobalt and zinc waves are merged. 
Iron and cobalt are estimated colorimetri- 
cally. If there is a large amount of zinc a 
third partial polarogram at lower sensitivity 
has to be taken to get the sum of the zinc 
and cobalt. 



Platinum, if present in the residue from 
the mercury distillation, is sometimes re- 
vealed in the polarogram by a spontaneous 
rise at the start and an anomalous dip or 
decrease in current after the lead wave. The 
dip usually begins to be evident at —0.9 
volt applied and reaches a minimum at 
— 1.10 volt applied. The addition of a drop 
or two of hydrochloric acid eliminates this 
dip but causes a catalytic hydrogen wave to 
start at —0.8 volt applied. Platinum may 
make it difficult or impossible to detennine 
the amount of zinc from the polarogram 
taken after adding pyridine. The terminal 
wave is shifted from —1.48 volts to — 1.2S 
volts by 0.01 mg of Pt in 3 ml. With 10 times 
as much platinum the terminal wave begins 
at -1.13 volts. 

Under the conditions chosen 10 micro- 
grams per 3 ml of any of the common 
metals Cu. Cd. Pb. Co. Xi, or Zn cause a 
wave of the ordei of 1 cm. Under the most 
favorable conditions 1 microgram of a metal 
may be detected, provided the metal does 
not occur in the reagents. If iron is the 
dominant component of the solution the 
sensitivity of the detection of lead may be 
greatly diminished because lead can be de- 
tected only in the most acid medium where 
the sensitivity that may be used is governed 
by the height of the initial iron wave. 

Other determinations. — Systematic pro- 
cedures have been adapted from colori- 
metric detenninations in the literature for 
the estimation of iron and cobalt in the 
solution that has been polarographed. The 
solution after electrolysis and separation of 

the electrolyzable metals may be tested 
systematically for such elements as molyb- 
denum by thiocyanate extraction, after 
which iron, titanium, and vanadium may 
be extracted as cupferrates and determined 
by -eparate colorimetric teste 


The possibility of using a small mercury 
cathode as a collector for minor amounts of 
electrolyzable metals has been explored and 
the procedure has been tested for a variety 
of applications. 

A novel combination of techniques based 
on the distillation of the mercury from the 
cathode and the polarographic and colori- 
metric examination of the residue has been 
devised and studied extensively in a rather 
limited range of applications. The pro- 
cedure appears to be capable of many 
further applications to a great variety of 
organic or inorganic substances. 


(1 Haight. G. P.. Jr.: McDuffie. R.: 
Spastk. G. W., and Furman. X. H. 
Madison Square Area Inf. Rep. A-1033. 
sect. 2. p. 1. Mar. 29, 1944. Revised 
Procedures. Cf. also Reports A-1045. 
2E. p. 1. Julv 4. 1944: A-1054. 2E. p. 1. 
Aug. 12.. 1944: A-1076. Mar. 20 ; 1945. 
2 LrxDELL, G. E. F.. and Hoffmax. J. I. 
Outlines of methods of chemical analysis. 
J. Wiley & Sons. Xew York. 193S. 

(3) Furmax. X. H.; Bricker. C. E.: and 

Whttesell. E. B. Ind. Eng. Chem.. 
.Anal. Ed. 14: 333. 1942. 

(4) Lixgaxr. J. J., and Kerllxger. H. Ind. 

Eng. Chem., Anal. Ed. 13: 77. 1941. 

ETHNOLOGY.— Self-torture in the Blood India?} sun dance. 1 
U. S. National Museum. 

John C. Ewers. 

When George Catlin, the artist and 
Indian enthusiast; published '1841) the 
first graphic account of the practice of 
self-torture in a Plains Indian ceremony, 
which he had witnessed in 1832. his de- 
scription was termed fantastic by D. D. 
Mitchell. Superintendent of Indian Affairs. 
Apparently shocked by Catlin" s vivid por- 
trayal of coolly premeditated self-sacrifice 
of human flesh and blood by participants 
in the Okipa ceremony of the Mandan, 

1 Received January 30, 194S. 

Mitchell declared. ''The scenes described 
by Catlin, existed almost entirely in the 
fertile imagination of that gentleman'' 
(Schoolcraft, vol. 3. p. 254; Catlin. vol. 1. 
pp. 157-177 . 

Catlin's description was substantiated. 
however, some years before Mitchell's 
accusations were made, by the Independent 
Mandan investigations of Prince Maximilian 
in 1833-34.. Maximilian, vol. 23. pp. 324- 
334). Catlin was defended strongly by the 
intelligent fur trader James Kipp, who had 

Mat V: 194S 

!•:■:■: ::~::^:- 5" 

: A>i 


:::: ~.:h Ci:lii "her. he ~l:ne5se:I :he 
ceremony *Kipp. pp. 436-438). 

Since the days of that historic contro- 
versy, the practice of self-torture in tribal 
sun-dance ceremonies has been reported, on 
reliable authority, as once characteristic 
of the Arapaho. Ankara, Assiniboin. Cana- 
dian Dakota. Cheyenne, Crow. Gros Ven- 
tres, Hidatsa. Oglala Dakota, Plains Cree. 
Plains Ojibway, Sarcee Siss eton Dakota, 
and the three Blackfoot tribes (Spier, pp. 
473 -475). Furthermore, brief accounts have 
been published of ceremonial self-torture, 
witnessed as early as 1805, among the 
Arikara and Hidatsa by the fur traders 
Pierre- Antoine Tabeau and Charles Mac- 
kenzie (Tabeau, pp. 191-193: Mackenzie, 
pp. 354-357 . 

It seems most probable that some Plains 
tribes practiced forms of self-torture in the 
period before first white contact. 

Among the Plains Indian tribes of the 
United States the practice of self-torture 
prohibited more than 60 years ago. 
This ban resulted from the combined op- 
position of missionaries and the civil and 
military branches of the Federal Govern- 
ment to such self-imposed cruelties, which 
tended to excite the Indians, to perpetuate 
both Indian-white and intertribal hostili- 
ties, and to make difficult the proce— 
civilization and Christianization of the 

Published descriptions of self-torture 
among these tribes have been primaiily of 
two kinds. Some were eye-witness accounts 
of interested but untrained white observers 
who not infrequently misinterpreted the 
purpose of the torture - eremony for 
"making braves." Others were based on the 
testimony of older Indians who had wit- 
nessed the tortures some years earlier but 
had not experienced torture themselves. 
Detailed case histories from the mouths of 
men who had submitted themselve- I 
torture are few and fragmentary in the 
extensive literature on the sun dance. 

Self-torture survived in the sun dances of 
the Blood and Xorth Blackfoot Indians of 
southern Alberta for a few years after its 
discontinuance among the Plains tribes of 
this country. In the course of field work on 
the Blood Reserve in September, 1947, the 

writer met two elderly full-bloods who had 
been tortured in the sun dance of their 
tribe. They were the last survivors of men 
of that tribe who had experienced this 
ordeal, and they were particularly desirous 
that their torture experiences should be 
recorded accurately. These narratives by 
Scraping White (now 81 years old) and 
Heavy Head (now 78). related to the writer 
through the interpreter Percy Creighton, 
provide new and significant information 
on the procedure of sell-torture in the 
Blood sun dance ceremony and its meaning 
to those who submitted to it. They help to 
round out the only published description 
of the Blood self-torture, that of the mis- 
sionary John McLean, who witnessed the 
ceremony prior to 1889. 

In the summer of 1889 the Blood medicine 
lodge was erected on the north bank of the 
Kootenay River, in southern Alberta. 
Three young men. Scraping White (then 
_ ears of age), Tough Bread (now de- 
ed), and Heavy Head (then 20 years 
old), presented themselves to be tortured. 

Sea _ T_ nite described his experience 

Three of us tortured ourselves in the sun 
dance that year — Tough Bread. Heavy Head, 
and I I was :he oldest of the three. 

I was on a war party to take horses from the 
Assiniboine when I made my vow to be tor- 
tured. Shortly before the sun went down, when 
we were in sight of the enemy camp, I turned 
to the sun and said. "I want good luck. Now 1 
z: '.: the r-r—y I — in: : : i-i-riir -i z: : h :rse 
and go home safely. Ill be tortured this coming 
sun dance As n as it was dark I went into 
the enemy camp and took two fast horses out 
of their corral without any of them knowing 
\ had good luck and reached home safely. 

Then I told my relatives of the vow I had 

made. Yellow Horn, an older relative, who had 

neer. '.'zt:'iz- :he Tirrure ~:e::re ::.i ne Pu: 

.• s~ e: " '. :zi : : .---:- : :.' I =h: h L: :k ; ::e: 

you." I made :he sweat h-Ige the very r_ex~ 

-'- : - 7 

Xot long after that the sun dance was held. 
The torture took place the day after the center 
pole was raised for the medicine lodge- I was 
the irs: :ne to be tortured. The torture began 
about noon. Old Yellow Horn cut my breasts 
~:h in irri. ;:: — he: . : :"_ . ::.s-::ei a ske~er 
thr-iu^h :h~ :u:s a: ea:h cress: These ^t _i :- 
were of serviceberry wood, flattened on both 



sides, thinned toward the ends but not sharp- 
ened, and about this long. [Scraping White 
indicated a distance of about 2 inches between 
his thumb and forefinger.] Then sinew was 
wrapped around the ends of the skewers and 
they were tied, each skewer to a 4-strand 
plaited rawhide rope. The two ropes were 
fastened at their far ends to the center pole at 
its forks. 

I stood up and Yellow Horn told me, "Now 
you walk up, put your arms around the center 
pole and pray. Tell s*un, now your vow is being 
fulfilled." I did just as he told me. Then I 
stepped back. Yellow Horn pulled hard on the 
rawhide ropes attached to the skewers. Then I 
danced. I didn't dance long before my flesh 
gave way and the skewers pulled out. Yellow 
Horn came to me and cut the skin that had 
broken. He trimmed it off even. Then he gave 
me the pieces of skin he had cut away and told 
me to take them and stick them in the ground 
at the base of the center pole, saying, "Now 
sun, I have completed my vow." 

Heavy Head's narrative of his torture 
experience was still more detailed : 

There were only two of us, Buffalo Teeth, 
my partner, and I. We went to war together to 
take horses. At Medicine Hat we found a small 
camp of Cree half-breeds. It was night when 
we saw their camp. It was moonlight. I looked 
up at the moon and prayed to it, "I shall be 
tortured at the sun dance if I have good luck 
and get home safely." Then I stole up to the 
camp and got one bay that was tied in front of 
a lodge without any of the enemy waking or 
seeing me. Buffalo Teeth took a roan. We 
started back to the Blood camp, traveled three 
days and three nights with no food other than 
a black rabbit. We got awfully weak and 

When I reached home I told my story to my 
father, Water Bull. The old man got up and 
sang his encouraging song. Then he told me, 
"My son, you have done something worth 
doing. You have made a vow that you will be 
tortured at the sun dance. You must do it this 
coming sun dance." 

A few days later I went out to the east point 
of Belly Butte to fast. While I fasted I dreamed 
that a sacred person came to me and gave me a 
drum and certain herbs to use for doctoring. 
Then I returned home. 

A short time after that the bands began to 
come together for the sun dance encampment. 
I prepared myself to go to an old man named 
Little Bear, a relative of mine, who had been 
through the torture himself, years before. I 

filled my pipe and took it to him. I gave him the 
pipe and a buckskin horse, and said, "Here is a 
horse for you. Keep this pipe too. I want you to 
look after me in the torture." When I gave him 
the pipe he put it down and went over to the 
next lodge. There were two old men there, Green 
Grass Bull and Red Bead. These men were not 
related to me, but they were both older than 
Little Bear, and both had been through the 
torture. Little Bear asked them to come to his 
lodge, to take my pipe and pray for me. After 
they prayed, they told me not to take any food 
or water the day I was to be tortured. 

The day before the torture I ate or drank 
nothing. Next day I ate or drank nothing until 
after the torture. However, the three old men 
gave me some sagebrush to chew. 

I was the last of the three Blood Indians to 
undergo the torture that day. Scraping White, 
who was the oldest, was first. Then Tough 
Bread, then I. I was the youngest. Inside the 
medicine lodge, on the west side of the center 
pole and north of the weather dancer's arbor, a 
shelter was built of sticks like a sweat lodge, 
covered with willow leaves. I went in there be- 
fore noon of the day of the torture. I was laid on 
my back with my head pointed north. I was 
barefoot, and wore only a breechcloth made 
from a small, red, trade shawl purchased from 
the Hudson's Bay Company. There was a little 
bowl of white paint and another of black paint 
nearby. The three old men painted four black 
dots, one below the other, under each of my 
eyes. This was called "tear paint." If I cried 
the tears would run down there. Then they 
painted a double row of six black dots on each 
arm. They painted the symbol of the moon, 
points up, on my forehead in black. On the out- 
side of each of my legs they painted a double 
row of six black dots. The rest of my body was 
painted white, also my face. They took some of 
the broad-leafed sagebrush from the ground 
inside the sweat lodge and bound it together, 
placed a wreath of it around my head, and 
bands of it around each wrist and ankle. 

I was taken from the sweat lodge and laid 
upon a blanket on the ground at the north side 
of the center pole with my head to the north, 
my feet toward the center pole. Other people 
were told to keep back away from me. Then an 
old man named Low Horn was brought for- 
ward. He counted four of his coups. The three 
old men, Little Bear, Red Bead, and Green 
Grass Bull, held me — one at each arm, and one 
at my head. Red Bead took a sharp, iron ar- 
rowhead in his hand, and asked me, "How do 
you want me to cut them? Thick or thin?" I 
said, "Thin." (I learned later that this question 

May 15, 1948 ewers: self-torture in blood indian sun dance 


Fig. 1. — Scraping White (left) and Heavy Head (right), two men who were 
tortured in the Blood sun dance of 1889. 

Fig. 2. — An act of self-torture in the Blood sun dance of 1891. Photograph by R. N. Wilson. 



was always asked of the man undergoing the 
torture before his breasts were pierced, and the 
one doing the cutting always did just the 
opposite of the young man's request. So when 
I said "thin," Red Bead knew to make his in- 
cisions deep). 2 Red Bead gave four of his own 
war coups. He made no prayer. Then he pierced 
my breasts with the sharp arrowhead and in- 
serted a serviceberry stick through each breast. 
The sticks were not sharp but flattened at the 
ends. The other two men held my arms as he 
cut and inserted the sticks. Blood flowed down 
my chest and legs over the white paint. Then 
Red Bead pressed the sticks against my body 
with his hands. They turned me around to face 
the sun and pierced my back. To the skewers 
on my back they hung an imitation shield, not 
so heavy as a war shield. The shield had 
feathers on it, but I don't remember how it was 
painted. It belonged to a man named Peninsula. 
The ropes were brought out from the center 
pole and tied to the skewers in my breasts — - 
right side first, then left side. Red Bead then 
grabbed the ropes and jerked them hard twice. 
Then he told me, "Now you go to the center 
pole and pray that your vow will come true." 
I walked up there. I knew I was supposed to 
pretend to cry. But oh! I really cried. It hurt 
so much. Coming back from the center pole I 
was shouting. Then, before I started to dance, 
I jerked the shield off my back. 

1 leaned back and began dancing, facing the 
center pole. It felt just like the center pole was 
pulling me toward it. I began to dance from the 
west toward the doorway of the sun lodge and 
back. Then, when the skewers did not break 
loose, the old men realized that the incisions 
had been made too deep. Red Bead came up 
and cut the outside of the incisions again so 
they would break loose. As I started dancing 
again the left side gave way and I had to con- 
tinue dancing with only my right side holding. 
An old man named Strangling Wolf jumped up 
from the crowd and came toward me shouting. 
He called out four coups he had counted and 
jumped on me. The last rope gave way and I 
fell to the ground. 

The three old men came to me and cut the 
rough pieces of flesh hanging from my breasts 
off even. They told me to take this flesh that 
had been trimmed off, and the sagebrush from 
my head, wrists, and ankles, and place them at 
the base of the center pole. I did as they told 

2 Jenness, pp. 54-5, reported the same con- 
trariness of action on the part of the Sarcee 
surgeon when the suppliant pleaded for a "thin" 

Then I took my robe and walked out of the 
medicine lodge alone. I went to a lonely place 
and fasted for a night. I wanted to dream. But 
I couldn't sleep at all because of the pain. At 
sunrise I prayed to the sun. 

Some time after that I saw a man approach- 
ing on horseback. He said, "I'm going to take 
you home right away." He took me up behind 
him on his horse and rode me slowly back to 
camp. My breasts were swollen and hurt. The 
rider's name was Red Crane. He told me of a 
mix-up that took place at the sun dance over 
horses stolen from the Gros Ventres. 

When I got to my lodge, my mother gave me 
something to eat. She and my father told me 
what had happened at the sun dance gathering 
— a mix-up between the Mounted Police and 
Indians. 3 I had to stay in the lodge several 
days. My breasts were so swollen I could 
hardly move. Indian doctors used herb medi- 
cines to take the swelling away and cure my 
wounds. 4 

The common elements in these two ac- 
counts reveal the pattern of the self-torture 
experience among the Blood Indians more 
clearly and completely than the brief de- 
scription of the ordeal previously published 
by McLean. This experience was initiated 
by a warrior through a vow to the celestial 
deities, sun or moon (the latter in Blackfoot 
belief was sun's wife), very shortly before 
the man exposed himself to danger. The 
vow was a simple, direct appeal to the 
deity for protection and success in the im- 
mediate, hazardous undertaking. In return 
for such aid the petitioner promised to make 
the self-torture sacrifice in the sun dance 
lodge of his people. On return home after 
the successful exploit the pledger made 
known his vow to his relatives. They helped 
him to obtain the services of one or more 
older men who had been through the ordeal 
and were qualified, therefore, to instruct and 
care for the young man in the ceremonial 
fulfillment of his vow. The public cere- 

3 In the summer of 1889, the Mounted Police 
sought to apprehend Calf Robe, a Blood horse 
thief, who sought sanctuary in the medicine lodge 
of his people. The Indians overpowered the police 
and set Calf Robe free, but there was no blood- 
shed (Steele, pp. 262-265). This incident serves 
to verify the year of Scraping White's and Heavy 
Head's torture. 

4 These narratives follow as closely as possible 
the words of the interpreter. In a few places de- 
tails obtained during further questioning of the 
informants have been inserted. 

May 15, 1948 ewers: self-torture in blood Indian sun dance 


mony took place in the sun-dance lodge, 
about midday of the day following the 
erection of the center pole. One man was 
tortured at a time. Each young man's 
experienced helper or helpers prepared him 
for the ordeal, pierced his breasts, inserted 
the wooden skewers, and attached the 
thongs leading from the crotch of the center 
pole. They guided his actions by telling him 
first to embrace the center pole and pray 
for successful fulfillment of his vow, watched 
him closely until he freed himself from the 
ropes, trimmed off the ragged edges of 
flesh from his breasts and instructed him 
to place them at the base of the center 
pole as an offering to the sun. With this 
act the vow was fulfilled. 

Both informants stated that the helpers 
were always men well advanced in years, 
rather than men who had been through the 
torture only a few years earlier. The num- 
ber of helpers depended primarily on the 
age of the victim and his relative's confi- 
dence in his ability to take the punishment. 
Younger men generally had more helpers 
to hold them as the incisions were made. 
Fasting was not considered obligatory. 
Scraping White said he was not required 
to abstain from food on the day prior to the 
torture. Scraping White's narrative omits 
mention of the painting of his body in prep- 
aration for the torture. However, Heavy 
Head asserted that all who underwent the 
torture were painted just as he had been. 

Neither Scraping White nor Heavy Head 
took any active part in the tortures at the 
sun dance of 1891, the last occurrence of 
self-torture in the Blood ceremony. They 
said that four men, Calf Tail, Buckskin 
Tom, Old Man Owl, and Takes Paint, all 
now deceased, were tortured in that year. 
R. N. Wilson, a trader on the Blood Re- 
serve, photographed these torture cere- 
monies. One of his photographs is repro- 
duced here, through the courtesy of Arch- 
deacon S. H. Middleton, principal of St. 
Paul's Residential School, on the Blood 
Reserve, owner of a print from the original 
glass plate negative. Percy Creighton be- 
lieved that the Indian shown in the act 
of torture was Takes Paint. 

The Blood Indians were the last Black- 
foot tribe, and probably the last tribe of 

Plains Indians to observe the self-torture 
ceremony. After their 1891 performance of 
the torture it was prohibited by the Indian 
Department and the Mounted Police. A 
year earlier the North Blackfoot had been 
persuaded to abandon torture (Ann. Rept., 
pp. 83-84). The Piegan eliminated torture 
from their sun dances at least 20 years 
earlier. Weasel Head (born about 1860), for 
many years a prominent weather dancer in 
the South Piegan sun dance, told the writer 
before his death in 1943, that he had never 
seen the torture performed in the sun dance 
of his tribe, although he recalled that as a 
youth he had seen older men who bore the 
scars of torture. Red Plume (born ante 
1850), informed Curtis that he had seen the 
torture rites in the Piegan sun dance only 
four times, thrice when a small boy and 
once when a young man. Each time a 
single man had submitted to the torture. 
Red Plume attributed the discontinuance 
of self-torture in the Piegan sun dance. to 
the warning of a North Piegan weather 
dancer that "they would die if they gave 
their bodies to the sun" (Curtis, vol. 6, p. 
55). Clark Wissler (p. 262) and Walter 
McClintock (p. 320) reported the persist- 
ence of this belief among the Piegan in the 
first decade of the present century. 

Wissler (p. 263) believed that self-torture 
had not become thoroughly adjusted to its 
place in the Piegan sun-dance ceremony at 
the time it was abandoned. On the other 
hand, the torture appears to have found 
much more favor among the neighboring 
Blood Indians. McLean (p. 236), stated 
that from two to five men underwent this 
torture every year in the Blood sun dance. 
Whereas the Piegan seem to have aban- 
doned torture as a result of native fear and 
distaste for the ceremony, the Blood con- 
tinued to practice it until they were com- 
pelled by Government authorities to give it 

Wissler (pp. 263-264) was inclined to 
credit a Piegan tradition that the Blackfoot 
tribes borrowed the torture ceremony from 
the Arapaho. There still exist among the 
Piegan vague but persistent traditions of a 
group of Blood Indians who sojourned for a 
number of years with the Arapaho in the 
early part of the nineteenth century. Chey- 



enne traditions, obtained from elderly men 
of the Southern Cheyenne nearly half a 
century ago by George Bird Grinnell and 
George E. Hyde, tell of a group of Gros 
Ventres and "Blackfeet" (division not in- 
dicated), who joined the Cheyenne and 
Arapaho in the Black Hills or on the Platte 
about the year 1826. According to one ver- 
sion of this tradition the " Blackfeet" 
returned north a few years later (Grinnell, 
vol. 1, pp. 39-40). Major Culbertson, who 
was married to a Blood woman, is reported 
to have found ten lodges of Blood Indians 
living with the Arapaho, when he attended 
the Fort Laramie Treaty Council in 1851. 
"They were unknown to him, and he did 
not learn how long they had been there 
or whether they ever returned' ' (Bradley, 
Book A, p. 184). These data, and the fact 
that the Piegan were relatively indifferent 
to the torture in later years, suggest the 
possibility that the torture feature was 
introduced among the Blackfoot tribes by 
Blood Indians, who may have borrowed it 
from the Arapaho no earlier than the second 
quarter of the nineteenth century. 

Spier (p. 491) regarded the self-torture as 
a nonessential element in the sun-dance 
ceremony of most Plains tribes. Our evi- 
dence certainly supports this conclusion in- 
sofar as the Blood sun dance is concerned. 
Both the torture pledger and his helpers 
played no necessary part in the sun-dance 
ceremony. They entered the medicine lodge 
for the sole purpose of fulfilling the pledg- 
er's vow. The Blackfoot tribes have con- 
tinued their annual sun dances to the pres- 
ent day, with no apparent sense of loss of 
any essential feature. Now as formerly the 
sun dance centers about the elaborate ritual 
prescribed for the fulfillment of the vow of 
the medicine woman who pledged the cere- 
mony. Her objective was reached with the 
completion of the medicine lodge, the day 
before the tortures took place. 

The Blackfoot tribes regarded the tor- 
ture, such as was endured by Scraping 
White and Heavy Head, as the most dan- 
gerous and severe form of physical sacrifice 
to the sun. The mutilation of the body by 
offerings of a finger or bits of flesh from the 
arms and legs were considered lesser ordeals 
(Wissler, pp. 263-265). Scraping White 

showed the writer scars on his legs resulting 
from the sacrifice of pieces of flesh to the 
sun prior to the year of his torture experi- 
ence. N evert heless, the belief of the Piegan 
that men who submitted to the torture 
would not live long after they had given 
their bodies to the sun, appears to have 
been based upon religious fear rather than 
the life histories of men who had been tor- 
tured. Several of our elderly South Piegan 
informants recalled having seen older men 
of the tribe who bore the scars of the torture. 
The severity of the torture varied with 
the depth of the incisions. Older men 
watched the suppliant carefully, and did 
not permit the torture to be prolonged in- 
definitely. Even in the case of Heavy Head, 
whose experience was described by other 
elderly Blood Indians as the most severe 
punishment they had witnessed in sun- 
dance tortures, the performance lasted 
only a few minutes. His narrative indicates 
the pain and nervous shock must have been 
intense. But in the majority of cases it 
probably did no permanent damage to the 
individual. Although Scraping White and 
Heavy Head still bear the scars of their 
torture, these men appear to be in fair 
health, active of mind and body, nearly 
six decades after they expiated their 
vows to the sun in the Blood medicine lodge 
in the summer of 1889. 


Annual Report of Department of Indian Af- 
fairs for 1891. Ottawa, Canada, 1892. 

Bradley Manuscript, in Montana Historical 
Society Library, Helena. 

Catlin, George. Letters and notes of the man- 
ners, customs and condition of the North 
American Indians, 2 vols. London, 1841. 

Curtis, Edward S. The North American 
Indian, 6. Norwood, Mass., 1911. 

Grinnell, George Bird. The Cheyenne In- 
dians, 2 vols. New Haven, 1923. 

Jenness, Diamond. The Sarcee Indians of 
Alberta. Canada Dept. Mines and Re- 
sources. Bull. 98, Anthropological Series 
23. Ottawa, 1938. 

Kipp, James. On the accuracy of Catlin's ac- 
count of the Mandan ceremonies. Ann. 
Rept. Board Regents Smithsonian Insti- 
tution for 1872. Washington, 1873. 

McClintock, Walter. The Old North Trail. 
London, 1910. 

McLean, John. The Blackfoot sun dance. 
Proc. Can. Inst., ser. 3, 6. Toronto, 1888. 

May 15, 1948 

dunkle: upper cretaceous selachians 


Mackenzie, Charles. The Missouri In- 
dians, a narrative of four trading expedi- 
tions to the Missouri, 180J+-1806. In L. 
R. Masson, "Les Bourgeois de la Com- 
pagnie du Nord-Ouest." Quebec, 1889. 

Maximilian (Prince of Wied-Neuwied). 
Travels in the interior of North America. 
Early Western Travels Edition, edited by 
Jteuben Gold Thwaites, vols. 22-24. 
Cleveland, 1906. 

Schoolcraft, Henry R. Information respect- 
ing the history, condition and prospects of 
the Indian tribes of the United States. 6 vols. 
Philadelphia, 1851-7. 

Spier, Leslie. The sun dance of the Plains 
Indians. Amer. Mus. Nat. Hist. An- 
throp. Papers, 16 (pt. 7). 1921. 

Steele, S. B. Forty years in Canada. Lon- 
don, 1915. 

Tabeau's Narrative of LoisePs Expedition to 
the Upper Missouri. Edited by Annie H. 
Abel. Norman, Okla., 1939. 

Wissler, Clark. The sun dance of the Black- 
foot Indians. Amer. Mus. Nat. Hist. 
Anthrop. Papers, 16 (pt. 3). 1918. 

PALEONTOLOGY. — On two previously unreported selachians from the Upper Cre- 
taceous of North America. 1 David H. Dunkle, U. 8. National Museum. 
(Communicated by C. Lewis Gazin.) 

Recently, the U. S. Geological Survey 
transferred to the National Museum two 
interesting toothlike fossils obtained by Dr. 
L. W. Stephenson during the course of his 
prolonged geological investigations on the 
Cretaceous formations of Texas. These 
specimens have proved to be examples of 
the structures currently interpreted as 
rostral teeth of pristid sharks. They are as- 
signable to the genera Onchopristis Stromer 
and Schizorhiza Weiler, both of which were 
founded on materials from various upper 
Cretaceous horizons and localities in North 
Africa. In the Western Hemisphere, Oncho- 
pristis has not heretofore been known and 
Schizorhiza only from a single, doubtful re- 
port from the upper Senonian of Chile 
(Wetzel, 1930; and Weiler, 1930). 

The definitive record of the fossil Pristi- 
dae in North America is confined to the iso- 
lated rostral armament of Pristis from the 
Cretaceous and a number of Tertiary hori- 
zons. In addition, during recent years the 
problematical genus Ischyrhiza has been in- 
cluded here. In the Eastern Hemisphere no 
less than 12 genera of fossil pristids are 
recognized. The majority of these are 
founded on isolated rostral teeth and their 
assignment to the family Pristidae, often- 
times doubtfully, has been based on the na- 
ture of their insertion onto the rostral carti- 
lages as shown by gross morphological 
features and by histological considerations. 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received January 
30, 1948. 

The present two specimens can not lend 
themselves to histological preparation with- 
out destruction, and little of general mor- 
phological and systematic information can 
be deduced from them. However, the new 
occurrences in essential stratigraphic con- 
temporaneity with their North African 
genotypes seems of sufficient interest to 
warrant the following brief description. 

The illustrations accompanying this note 
have been prepared by Mrs. Elinor Strom- 
berg, scientific illustrator of the U. S. Geo- 
logical Survey. 

Onchopristis cf. numidus (Haug) 
A single unassociated tooth (U.S.N. M. no. 
17088), although with only one side exposed 
and lacking the distal extremity, exhibits well 
the characteristics of the genus as defined by 
Stromer (1917). 

The structure is strongly compressed, dorso- 
ventrally, and is composed of a short inserted 
base and an exposed, enamel-covered crown. 
The line of demarcation between these two 
parts is prominent, being remarked by a con- 
striction of the base adjacent to the proximal 
limit of the enamel. This latter shows a border 
arched in gentle convexity toward the base and 
extended obliquely across the long axis of the 
element. The preserved portion of the crown 
possesses one posterior barb. Both anterior 
and posterior margins of the crown form sharp 
cutting edges except at the proximal reen- 
trance of the barb where the edge is bluntly 
rounded. The enamel on the slightly convex, 
exposed surface of the tooth is checked both 



Fig. 1. — Onchopristis cf. numidus (Haug) 
(U.S.N.M. no. 17088). Rostral tooth from the 
Woodbine formation of Texas. Approx. X4. 

longitudinally and transversely as a result of 
weathering but exhibits no observable orna- 
mentation. The base, in dorsoventral aspect, is 
expanded and is marked by a few regularly 
spaced furrows, which are extended in the 
direction of the longest axis of the tooth. 

Measured in relation to the longest axis, the 
specimen has the following dimensions: Total 
preserved length, 13.7 mm; greatest length of 
base, 4.5 mm; maximum width of base, 8.4 
mm; width of crown immediately below barb, 
6.0 mm; and greatest thickness of crown on 
fractured distal exposure, 2.4 mm. 

Remarks. — The systematic history of On- 
chopristis numidus (Haug) has been completely 
summarized by Stromer (1917). The present 
example from Texas differs from the North 
African specimens in the shortness of that 
portion of the enameled crown proximal to the 
barb. Stromer (1917, 1925, and 1927), however, 
in his discussions on this sawfish, has demon- 
strated wide variation between individual ros- 
tral teeth and has described examples with 
multiple posterior barbs. In consequence there 
seems to be no immediate basis for erertimi 
additional species within the genus. 

Geological horizon and locality. — Collected 
from the Upper Cretaceous (Cenomanian) in 
the Lewisville member of the Woodbine for- 
mation, approximately 2.4 miles southwest of 
Lewisville, Denton County, Tex., by L. W. 
Stephenson, October 25, 1945. 

Distribution. — All the North African occur- 
rences of Onchopristis are considered by 
Stromer to be of Cenomanian Age. 

Schizorhiza cf. weileri Serra 
To this form is assigned a small imperfectly 
preserved tooth (U.S.N.M. no. 17087). The 
complete outline of one side of the specimen is 
clearly impressed in the limy sandstone matrix, 
although almost all the tooth tissue from that 
side has been lost. The opposite side of the 
tooth is deeply imbedded in the matrix of the 
counterpart. As thus incompletely exposed it 
was necessary to crack the counterpart in two 
and examine the cross section of the imbedded 
part before any structural interpretation could 
be made. 


Fig. 2. — Schizorhiza cf. weileri Serra (U.S.N.M. no. 17087). Rostral tooth from the Escondido for- 
mation of Texas in (A) dorsoventral aspect as restored from a cast of the preserved impression and 
in (B) ideal transverse section. Approx. X4. 

May 15, 1948 

dunkle: upper cretaceous selachians 


It is assumed that this tooth projected later- 
ally from firm attachment along the edge of a 
rostrum. In orientation, therefore, the long 
axis of the tooth is a transverse dimension, and 
this breadth exceeds the shorter length by 
about three times. The element is moderately 
compressed, dorsoventrally, and the top and 
bottom halves are apparently symmetrical. 
The crown occupies approximately one-third 
the longest axial dimension, and from the 
shiny smoothness of its impression is indicated 
to have been enamelled. It projects directly out 
in the frontal plane of the tooth without dorsad 
or ventrad flexure; is arched very slightly to- 
ward the rear; and is basally emarginated at 
the anterior and posterior edges. The free lateral 
margins are sharp, straight, and converge to an 
obtuse distal point. The root, viewed either 
dorsally or ventrally, is seen to expand proxi- 
mally from a narrow distal neck. The part is 
divided into flat upper and lower lips by a 
frontal groove which extends outward nearly 
to the base of the crown. Three furrows trav- 
erse the outer surfaces of both upper and lower 
lips, parallel to the long axis of the element. 
These grooves are continued proximally as 
deep notches in the inner edges of the root 
lips. Thus, mesially, the lips are each produced 
into four acutely conical projections, of which 
the central two are the longest. 

Measured in relation to the longest axis, the 
tooth possesses the following dimensions: Total 
length 15.0 mm; length of crown, 4.0 mm; maxi- 
mum width of crown, 5.0 mm; maximum width 
of root, 6.0 mm; length of longer proximal 
rays, 5.5 mm; and depth of proximal divergence 
of root lips, 6 mm. 

Remarks. — The genus Schizorhiza is known 
only from isolated teeth and the genotype is 
the species stromeri (Weiler, 1930). In addition 
to some Egyptian specimens, Weiler included 
in his original description a number of frag- 
mentary teeth from several widely scattered 
localities (Di Stephano, 1919; Quaas, 1902; and 
Wetzel, 1930). This type series presented a 
range of structural variation from teeth as 
here described to examples which in gross 
features appear similar to Ischyrhiza, with long 
tapering, scimiterlike crowns and robust roots 
exhibiting only a shallow frontal groove. Ap- 
parently no one specimen among those was 
designated the type. Subsequently, Serra 
(1933) abandoned the name stromeri for all of 

the teeth of the type series except that pic- 
tured by Weiler in his Fig. la, PL II. In return, 
the specific name weileri was proposed for the 
tooth illustrated as Fig. 3a, PI. II, and a col- 
lection of identical specimens from Sofeggin, 

Of these two species, the present Texas speci- 
men is more closely comparable with weileri. It 
differs from it only in the more pronounced 
asymmetry of the crown profile in dorsoven- 
tral aspect and in exhibiting a greater size than 
reported by Serra. Provisional assignment to 
the latter species is therefore made. While 
possibly representing a new species, it does not 
seem advisable to establish a type on this, an 
incomplete tooth heretofore unrecorded from 
North America. A more critical evaluation will 
depend upon the acquisition of a series of the 
Texas specimens which permit qualitative 
studies and histological examination. 

Geological horizon and locality. — Collected 
from the Upper Cretaceous (Maestrichtian) in 
the Escondido formation, 2 miles east of Eagle 
Pass, Maverick County, Tex., by L. W. Ste- 
phenson, October 30, 1912. 

Distribution. — Weiler (1930) gave a Senonian 
Age to all the teeth from North Africa, Arabia, 
and Chile on which he based his definition of 
Schizorhiza stromeri (sensa lato). Serra (1933) 
considered the occurrence of S. iveileri as 


Haug, E. (Paleontologie, in Foureau, T.) 

Documents scientifiques de la mission sa* 

harienne 2. Paris, 1905. 
Quaas, A. Beitrag zur Kenntnis der Fauna cler 

obersten Kreidebildungen in der Libyschen 

Wiiste (Overwegischichten und Bldttertone) . 

Paleontographica 30 (3): 153-334. 1902. 
Serra, G. Di nuova specie di Schizorhiza del 

Maestrichtiano delta Tripolitania. Riv. 

Ital. Paleont. 11 (2-3): 103-107, pi. 3. 

Stephano, G. de. Osservazioni sul Cretaceo e 

sul Eocene del deserto Arabico e di Sibaiya, 

nella valle del Nilo. Boll. Comit. Geo!. 

Ital. 47: 1-39, pis. 1-11. 1919. 
Stromer, E. Die Sage des Pristiden Oncho- 

pristis numidus Haug sp., und iiber die 

Sagen der Sagehaie. Abh. Bayer. Akad. 

Wiss., math.-nat. Abt., 28 (8): 1-28, 1 pi. 

. Ergebnisse der Forschungsreisen Prof. 

E. Stromer in den Wusten Agyptens. II. 

Wirbeltier-Reste der Baharije-Stufe (un- 

terstes Cenoman). 8. Ein Skelettreste des 

Pristiden Onchopristis numidus Haug sp. 



VOL. 38, NO. 

Abh. Bayer. Akad. AYiss., math.-nat. Abt., 
30 (6): 11-12, 1 pi. 1925. 

— . Ergebnisse der Forschungsreisen Prof. 
K. Stromcr in den W listen Agyptens. II. 
Wirbeltier-Reste der Baharije-Stufe (wri- 
ter stes Cenoman). 9. Die Plagiostomen 
mil einem Anhang ilber kdno- und meso- 
zoische Riickenflossen-Stacheln von Elasmo- 
branchicrn. Abh. Bayer. Akad. Wiss., 
math.-nat. Abt., 31 (5): 1-64, 3 pis., 14 
figs. 1927. 

Weiler, W. (In Stromer, E., and W. Weiler). 

Beschreibung von Wirbeltierresten aus dent 
nubischen Sandsteine Oberagyptens und aus 
agyptischen Phosphaten nebst Bemerkungen 
\xber die Geologie der Umgegend von Ma- 
hamid in Oberagyptens. Abh. Bayer. 
Akad. Wiss., math.-nat. Abt., new ser., 7: 
1-42, 4 pis. 1930. 
Wetzel, W. Die Quiriquina-Schickten als 
Sediment und palaontologisches Archiv. 
Palaeontographica 73 (Fishes) : 94-97. 1930. 

ENTOMOLOGY. — Synoptic revision of the United States scarab beetles of the sub- 
family Dynastinae, No. 4: Tribes Oryctini (part), Dynastini, and Phileurini. 1 
Lawrence W. Saylor, California Academy of Sciences. 

This paper is the fourth in the series of 
my United States dynastine scarab beetle 
studies and completes the specific listings 
and notes. The fifth, and last, part will in- 
clude a complete classification of the tribes 
and genera, from the Nearctic standpoint. 

Genus Aphonus LeConte 

Aphonus LeConte, 1856, p. 21; Horn, 1882, p. 
122; LeConte and Horn, 1883, p. 259; Casey, 
1915, pp. 178, 210; Ritcher, 1944, p. 28, Cart- 
wright, 1944, p. 36. 

Podalgus (part) Burmeister, 1847, p. 117; Lacor- 
daire, 1856, p. 408. 

Aphonus (as now constituted) is limited to 
the Eastern United States and contains four 
valid species; the farthest west I have knowl- 
edge of the genus occurring is Texas, where 
brevicruris Cartwright was collected. 

The only character separating the adults of 
Aphonus from United States species of Cheiro- 
platys Hope is the trilobed (varying to sub- 
tridentate to even simply carinate in worn ex- 
amples) preapical carina of the clypeus. In the 
larvae, Ritcher has pointed out (1944) the very 
close similarity between Cheiroplatys pyriformis 
LeConte and Aphonus castaneus (Melsheimer), 
the only real difference being that the first 
antenna! segment in the latter is bare of setae, 
and the other two key characters being those 
of degree only (relative distance between lobes 
of the thoracic spiracle, respiratory plate, and 
width of the head capsule). In the adults, the 
peculiar preapical carina and the rather odd 
apical, front tibial tooth, as well as external 
facies and proportions in general, immedi- 
ately disclose the close affinity of the two 
groups at present called Aphonus and Cheiro- 

1 Received September 3, 1947. 

platys. Indeed, I have relatively fresh speci- 
mens of A. castaneus (from Massachusetts and 
New Hampshire) in which the preapical clypeal 
carina is distinctly bidentate, so that the only 
character for the retention of the name 
Aphonus as a valid genus is gone. 

How r ever, even though I feel that Aphonus 
must eventually be considered a synonym of 
the earlier described Cheiroplatys I am unable 
definitely to synonymize the two genera until 
I can review the genotypes; I desire also to 
dissect carefully the mouthparts of many of 
the Neotropical and Australian forms of the 

The sexes of Aphonus are rather similar in 
most characters, but in the male the last ab- 
dominal sternite is shorter and is feebly but 
distinctly emarginate apically, whereas the last 
abdominal in the female is longer and the apex 
is evenly rounded. 

I have been unable to construct a satisfac- 
tory key to the species based on nonvariable 
external differences, so that to place the species 
properly it is necessary to make genital dissec- 
tions. The only general statement that can be 
made as to the external facies is that castaneus 
is most frequently rufous, averages 10 to 11 
mm in length, and is more robust; whereas 
densicauda and tridentata are both more 
elongate, and the former averages 13 mm in 
length and the latter 15 mm. Obviously, such 
generalities are interesting but of little assist- 
ance in actually separating closely allied and 
variable species. According to Cartwright's 
description of brevicruris (the unique type of 
which I have not seen) the proportions of the 
heavy, short tibia and femora wall readily sep- 
arate this species from all other described 

May 15, 1948 saylor: synoptic revision of subfamily dynastinae 


Aphonus castaneus (Melsheimer) 
Fig. 1, k, m 

Bothynus castaneus Melsheimer, 1856, p. 138; 

LeConte, 1856, p. 22. 
Podalgus obesus Burmeister, 1847, p. 119; Arrow, 

1909, p. 341. 
Aphonus castaneus (Melsheimer) Casey, 1915, p. 

220; Sims, 1934, p. 334 (larvae); Johnson, 

1942, p. 79; Ritcher, 1944, p. 30, pis. 2-5 

Aphonus cubiformis Casey, 1915, p. 221. 
Aphonus saginatus Casey, 1915, p. 220. 
Aphonus trapezicollis Casey, 1915, p. 219. 

All examples of this small species that I have 
examined vary from rufocastaneous to piceo- 
castaneous in color, and from 8 to 13 mm in 
length. The species ranges generally along the 
East Coast from Maine south through the 
Carolinas, Georgia, and Alabama. Johnson 
mentions finding numerous larvae on the sur- 
face of a Connecticut golf course during a gen- 
tle July rain, and these larvae retreated into 
the sod when the sun reappeared. Sims records 
the larvae as common in the turf of the coastal 
plains golf courses and in sandy soil generally. 

Aphonus densicauda Casev 
Fig. l,h,l 

Aphonus densicauda Casey, 1915, p. 216; Ritcher, 
p. 31 (larvae). 

Described from Pennsylvania, and seen also 
from New Hampshire, south to Georgia and 
Kentucky and west to Iowa; will probably be 
found to have a much wider distribution than 
indicated by available specimens. Ritcher re- 
cords it as fairly common in Kentucky where 
full-grown larvae may be found in pastureland, 
in or just beneath the sod, from November to 
May, and pupation occurs late in May or 
early in June; he found adults in the soil 
throughout the year. 

The color is usually piceous or piceocastane- 
ous, varying to rufous, as do all species of the 
genus, and the length averages 13 mm. The 
adults are hard to separate from typically 
black tridentatus other than on genital charac- 
ters and the slightly larger average size of the 
latter (15 mm.) ; the larvae are also very similar 
but distinct according to Ritcher (1944). 

Aphonus tridentatus (Say) 
Fig. 1,/, i,j, n 

Scarabaeus tridentatus Say, 1823, p. 209. 
Bothynus variolosus LeConte, 1848, p. 88 (new 

Aphonus tridentatus (Say) Horn, 1882, p. 122; 
Casey, 1915, p. 215; Ritcher 1944, p. 33 (lar- 
vae); Arrow, 1937, p. 42 (additional refs.). 

Aphonus aterrimus Casey, 1915, p. 216. 

Aphonus congestus Casey, 1915, p. 218. 

Aphonus elongatus Casey, 1915 ; p. 215. 

Aphonus f rater LeConte, 1856, p. 22. 

Aphonus hydropicus LeConte, 1856, p. 22. 

Aphonus ingens Casey, 1924, p. 334. 

Aphonus modulatus Casey, 1915, p. 219. 

Aphonus politus Casey, 1915, p. 218. 

Aphonus scutellaris Casey, 1924, p. 335. " 

Usually piceous, this largest United States 
species of the genus varies to entirely rufous, 
especially in specimens from Florida, and these 
latter are the variolosus of LeConte; the Florida 
specimens are often 2-3 mm smaller than the 
more northern specimens and superficially look 
different, but the genitalia and all essential di- 
agnostic characters are identical and I have no 
doubt of the correctness of the synonymy. I 
have seen specimens from Michigan, Indiana, 
Illinois, south through Georgia, South Caro- 
lina, and Florida; also recorded from New 
York and Wisconsin. Cartwright has taken 
numbers at Clemson, S. C, from March 
through July. Ritcher says that the larvae are 
found in woodland loam; collected by Yeager 
from "forest duff" in Michigan. 

Aphonus brevicruris Cartwright 
Fig. 1, e 

Aphonus brevicruris Cartwright, 1944, p. 36, pi. 1. 
fig- 5. 

Described from a unique male collected at 
Austwell, Tex., May 20, 1941 (Goodpaster 
collector), and not taken since to my knowl- 
edge. I have not seen the type, and the informa- 
tion here is reworded and taken from Cart- 
wright's paper: Easily separable from all other 
United States species by the proportions of 
the hind legs: the femur is three-fifths as wide 
as it is long, the tibia is shorter than the femur, 
and the tibial apex is widely flared and more 
than half as wide at apex as the full tibial 
length; in all other United States species the 
hind femur is only half&s wide as long, the hind 
tibia and femur are subequal in length and the 
hind tibial apex is flared but at most is one- 
third as wide at apex as the length. 

Tribe Dynastini 

Some of the largest and heaviest insects 
in the world occur in this tribe, including 



the well-known Dynastes hercules of the 
American Tropics. Arrow (1937) lists only 
21 genera in the tribe from the world, many 
of these genera being monobasic. The essen- 
tial character of the enlarged male forelegs 
is not too well shown in our United States 
species, but in our relatively common Golofa 
Hope and such Megasoma as elephas the 
character is strongly indicated. In the 
Colombian Golofa porteri Hope the front 
legs in the male are as long as the entire 
body and exactly twice as long as the fore- 
legs of the females. Bates (1889) reports the 
immense Megasoma elephas (Fabricius) as 
feeding in numbers on ripe mangoes in 
Panama, and my father has collected num- 
bers around street lights in central Panama, 
where the large lumbering insects often fly 
into the faces of passersby and occasionally 
badly scratch or at least scare them! Since a 
large male specimen weighs nearly half a 
pound, the bruises and abrasions that could 
be occasioned by such a specimen flying 
into a person's face is easily imaginable. 

We have only two genera in the United 


First segment of hind tarsus sharp on outer side 
but not really extended into a long spine (length 
of segment on outer side exclusive of apical 
movable setae only one-third to one-fifth 
longer than length on inner side); prosternal 
spine high between front coxae (as "tall" as its 
own width across base) and always either 
densely hairy or at least hairy or setose api- 
cally on posterior side; surface glabrous and 
usually gray, speckled with piceous spots 
(rarely unicolorous in some females) ; male with 
front thoracic angles normal, that is, not sinu- 
ous (southeast and southern United States and 
Mexico) Dynastes Kirby 

First segment of hind tarsus with a long distinct 
apical spine (length of segment on outside, in- 
cluding spine, one-half to three-fourths times 
longer than length on inner side); prosternal 
spine much shorter than coxal length (two- 
thirds as "tall" as its own width across base) 
and always quite glabrous on external face; 
surface always hairy (velvety) and unicolorous 
piceous; male with each front thoracic horn 

strongly sinuous (Arizona and Mexico) 

, Megasoma Kirby 

Genus Dynastes Kirby 

Dynastes Kirby, 1825, p. 568; Burmeister, 1847, 
p. 256; Lacordaire, 1856, p. 444; Casey, 1915, 
p. 258; Arrow, 1937 (many references given), 
p. 95; Ritcher, 1944, p. 39 (larvae). 

Our species have been variously listed or de- 
scribed in the genera Scarabaeus, Geotrupes, 
Xylotrupes, and others by the older authors and 
such references are readily available in Arrow 
(1937) and Burmeister (1847). 

Arrow lists 13 species of these so-called 
"rhinoceros beetles" as valid in his 1937 cata- 
logue, these occurring in India, the Philippines, 
Java, Burma, Borneo, Nigeria, the Congo, and 
the Americas. I seriously doubt that all these 
could possibly be congeneric. At any rate, six 
species are listed from the Americas, and two 
of these are supposed to occur in the United 
States. The large and well-known hercules ap- 
parently has not been taken yet north of Gua- 
temala, except possibly in quarantine intercep- 
tions, and it appears to be replaced in Mexico 
by the much smaller so-called hyllus of Chevro- 
lat. I have seen specimens of the latter species 
from central and southern Mexico and am 
entirely unable to separate them either on 
genital or external characters from our common 
tityus (Linnaeus). 

Our Arizona Dynastes are usually called 
granti Horn; the male genital characters are 
identical with those of tityus and the only dif- 
ferences can be summed up in the following 
(the thoracic horn is measured with a microme- 
ter scale in a direct line between the laterobasal 
denticles at each side of the horn base, and the 
horn apex): 

Male thoracic horn distinctly 6-8 times longer 
than scutellar length (19-23 mm/3 mm); this 
horn also much broader basally through the 
denticles; horn on head long, with a very dis- 
tinct and large, preapical tooth on the dorsal 

side (Arizona) granti Horn 

Male thoracic horn 3-5 times longer than scutel- 
lar length (7-12 mm/2.5 mm); horn of head 
usually faintly notched dorsally or entirely 
smooth, never with a dorsal preapical tooth. 

Eastern United States and Mexico 

tityus (Linnaeus) 

As anyone who has worked to any extent in 
the Dynastini knows, these characters as listed 
above are highly variable in a group where ex- 
ceptionally dimorphic forms are the rule rather 
than the exception, and I am not at all sure of 
the validity of granti. We have an exact coun- 
terpart of this in the related and well-known 
Golofa imperialis Thomson and pizarro Hope 
where the males are unusually variable and the 
thoracic horn in male majors is unusually long 
and toothed within, varying through all de- 
grees to the male minors, in which the horn is 

May 15, 1948 saylor: synoptic revision of subfamily dynastinae 


the merest sort of a knob with a smooth inner 
surface. Until such time as exact intermediate 
specimens can be collected between Arizona 

and the more typically eastern tityus, it appears 
best to retain the name granti as a weak sub- 
species of tityus. 

Fig. 1. — a, Phileurus truncalus (Beauvois): Male genitalia; b, Archophileurus cribrosus (LeConte): 
Male genitalia; c, Phileurus Hiatus LeConte: Male genitalia; d, Phileurus castaneus Haldeman: Male 
genitalia; e, Aphonus brevicruris Cartwright: Male genitalia;/, Aphonus tridentatus (Say): Head of female, 
front view; g, Dynastes tityus (Linnaeus): Male genitalia; h, Aphonus densicauda Casey: Male genitalia; 
i, Aphonus tridentatus (Say): Male genitalia, from Jacksonville, Fla.;y, Aphonus tridentatus (Say): Male 
genitalia, from South Carolina; k, Aphonus castaneus (Melsheimer); Male genitalia; from Rhode Island; 
I, Aphonus densicauda Casey: Female genitalia; m, Aphonus castaneus (Melsheimer): Female genitalia; n, 
Aphonus tridentatus (Say): Female genitalia. 



VOL. 38, NO. 5 

Dynastes tityus (Linnaeus) 
Fig. 1, g 

Scarabaeus tityus Linnaeus, 1763, p. 391. 
Scarabaeus marianus Linnaeus, 1767, p. 549. 
Scarabaeus pennsylvanicus DeGeer, 1774, p. 308. 
Scarabaeus hyllus Chevrolat, 1843, p. 33 (new 

synonymy); Bates, 1888 (as Dynastes), p. 336; 

Duges, 1887 (as Dynastes), p. 137 (biology). 
Scarabaeus iphiclus (Panzer) Burmeister, 1847, 

p. 259. 
Dynastes tityus (Linnaeus) Burmeister, 1847, p. 

260; Lacordaire, 1856, p. 444; Arrow, 1937, 

p. 98 (many references); Casey, 1915, p. 260; 

Hamilton, 1886, p. 112 (biology); Manee, 1915, 

p. 266 (biology); Ritcher, 1944, p. 39 (larvae). 
Dynastes corniger Sternberg, 1910, p. 26 (new 

sy no no my). 
Subspecies: Dynastes granti Horn, 1870, p. 78; 

Casey, 1915, p. 261;' Arrow, 1937, p. 97. 

This large and familiar species is widespread 
throughout the eastern United States from 
New York and Pennsylvania south through 
Florida, west to Arizona and south into Mexi- 
co, and possibly Guatemala. Varies greatly in 
size and color, especially in the females. The 
smallest specimen I have seen was 37 mm and 
the largest 74 mm; with the average about 55 
mm. The often asymmetrical (bilateral) colora- 
tion has been frequently noted, especially in 
the females, and a good description is given by 
Ritcher (1944) of this variation in adults he 
collected in a single stump and very probably 
from a single parent: in 14 pupal cells (7 males, 
7 females), 8 individuals were spotted, 5 had 
one elytron spotted and the other of a solid 
dark mahogany color, and 1 was of a uniform 
dark mahogany color. The larvae are recorded 
as feeding in decaying wood of oaks, pines, 
chestnut, willows, wild cherries, black locust, 
and fruit trees such as peaches and apples. 
The adults feed on the sap of wounded trees as 
well as decaying fruit of peaches, plums, pears, 
and apples, and Casey claims that the adults 
have a characteristic odor that can be smelled 
for some distance, if the observer is downwind 
of a considerable number of individuals. 

D, granti Horn was described from Arizona, 
though tityus has also been recorded from this 
State. As stated above, I am not at all sure that 
the form is sufficiently distinct to warrant its 

Genus Megasoma Kirby 

Megasoma Kirby, 1825, p. 566; LeConte and 
Horn, 1883, p. 260; Casey, 1915, p. 261; Arrow, 
1937, p. 98 (other references). 

Megasominus Casey, 1915, p. 261; Arrow, 1937, 

p. 35. 
Megalosoma Burmeister, 1847, p. 273. 
Lycophontes Bruch, 1910, p. 73. 

Arrow lists eight species of this American 
genus, five of them from South America. Our 
only species in the United States is the ther- 
sites of LeConte, which occurs in Arizona and 
Lower California. Casey erected the genus 
Megasominus for this species, the essential dif- 
ferences between it and the very much larger 
Neotropical elephas (Fabricius) being the pres- 
ence of a large basal horn dorsally on the base 
of the male cephalic horn in the latter, and also 
the marked sexual dimorphism in the front 
legs of the two sexes of elephas (fore tibia of 
male noticeably elongated and curved, and not 
so in female); in thersites there is no trace of a 
dorsal tooth on the base of the large cephalic 
horn, nor is there any difference between the 
length of the front legs in the two sexes. At 
first glance, there is a marked difference be- 
tween male elephas and thersites: the former is 
very large (3| to 4| inches) and with a dense 
uniform clothing of short velvety pile, and a 
very large forward-projecting tooth on the base 
of the cephalic horn, and the mid-disc of the 
thorax is evenly convex without a central horn ; 
whereas in thersites the male is small (1| to \\ 
inches long), the cephalic horn has no basal 
tooth, the dorsal clothing of pile is slightly 
longer, less velvety, and much less uniformly 
placed, and the mid-disc of the thorax has a 
narrow, semierect horn, which is slightly bi- 
furcate apically. However, in the females of the 
two species, the only essential difference be- 
sides size is that the thorax base in elephas is 
strongly margined, as opposed to the non- 
margined base of thersites, but all other essen- 
tial characters are so closely similar that it is 
necessary to treat the two species as congeneric. 

Megasoma thersites LeConte 

Megasoma thersites LeConte, p. 336. 
Megasominus thersites (LeConte) Casey, 1915, p. 

This uncommon species is usually confined 
to Lower California, but I have a specimen 
taken in the "Coyote Mts., Arizona, August, 
3,500 ft. elevation." Ross and Bohart collected 
the species at San Venancio in Lower Cali- 
fornia on October 8, 1941. Easily separable in 
the male from all other LTnited States Dvnas- 

May 15, 1948 saylor: synoptic revision of subfamily dynastixae 


tini by the combination of the strongly bi- 
furcate clypeal horn, and the sharp tooth of 
each front thoracic angle, as well as the mod- 
erate to short, narrow, erect, weakly bifurcate 
horn of the mid-disc of the thorax. The female 
is readily separable from female Strategus by 
the widely separated front teeth on the apex 
of the clypeus (one at each side angle) as well 
as the sharp, bidentate mandibles and the non- 
margined center base of the thorax. The female 
ther sites somewhat resembles a female Apho- 
nides dunnianus, but the clypeus there is uni- 
dentate at apex and the mandibles are equally 
rounded and not at all toothed. The life his- 
tory is apparently unknown. 

It is very surprising to me that the male 
aedeagus of this species is inseparable in form 
from that of male Dynastes tityus. but such ap- 
pears to be the case, based on my own careful 
dissections; throughout dynastines generally, 
the characters of the male genitalia appear to 
be specific within narrow boundaries of varia- 

Tribe Phileurini 

This tribe is the most aberrant of the sub- 
family Dynastinae, as the labial palpi are 
inserted on the underside of the ligular 
plate instead of at the sides, and the gener- 
ally black color and the depressed (majority 
of species) dorsal surface is suggestive of the 
Passalidae. Numerous genera and species 
are described and the tribe is world-wide; 
our American (Neotropical) species and 
genera are very poorly and inadequately 

The two sexes are not well differentiated 
externally in this tribe. The only obvious 
external sexual difference is that the male 
sixth abdominal sternite (instead of being 
emarginate apically as in most other 
dynastine tribes) is subtruncate at apex, 
whereas the same sternite in the female is 
somewhat narrowly rounded, though at 
times it is practically impossible for even an 
experienced student of the group to be cer- 
tain of the sex unless he dissects the speci- 
men. This condition, however, does not hold 
good throughout the tribe, since males of 
the Neotropical Amblyodus Westwood show 
the typical emargination of the last sternite. 


1. Completely lacking any tubercles or horns on 

either head or thorax; clypeal base consisting 
of a wide carina which is obsolescent lat- 
erally; front not at all concave, but coarsely 
punctate; side of thorax exceptionally hairy 
(hairs really extend from beneath thoracic 
margin) ; elytra short and very coarsely crib- 
rate; apex of hind tibia slightly irregular 
but not really toothed; all tarsal segments 
short and subrectangular in shape; elytra 
"soldered" at sutures and wings reduced to 

mere vestiges 

Archophileurus cribrosus (LeConte) 

With tubercles or horns on either head or 
thorax or both, without transverse clypeal 
carina; front always strongly concave; elytra 
longer, depressed; hind tibial apex distinctly 
spinose, at least at sides; tarsal segments 
longer, basal segment of mid and hind legs 
with strong apical spine; elytra not soldered 
at suture and wings of normal length. 
(Phileurus) 3 

2. Size large (29-34 mnO ; clypeal horn very large, 

as long as exposed dorsal portion of head, 
each horn situated right at and on lateral 
margin of head; small canthus in front of 
eye (dorsal view) very obsolescent, not at all 
conspicuous. .Phileurus truncatus (Beauvois) 
Size much smaller (16-23 mm); clypeal horn 
small or represented by a tubercle, always 
much shorter than head length ; clypeal horn 
situated inside each lateral margin, and not 
at it; eye canthus moderate to strong 3 

3. FroDt tibia distinctly 4-dentate, the subapical 

and apical external teeth very narrowly sep- 
arated by a distinctly U-shaped incision; 

the sides of the "incision" parallel 

Phileurus castaneus Haldeman 

Front tibia tridentate, at most with the merest 
suggestion of a fourth tooth (near base if 
present I apical and subapical external tooth 
separated by a wide non-parallel-sided emar- 
gination Phileurus Hiatus LeConte 

Genus Archophileurus Kolbe 
Archophileurus Kolbe, 1910, p. 334; Casey, 1915, 
p. 271; Arrow, 1937, p. 38; Cazier, 1399, p. 170. 

Arrow in his 1937 catalogue lists a number of 
American species in this genus, some of these 
however actually being synonyms, but the 
group is not well enough known to definitely 
list them as such at the present time. Our 
single United States species also occurs in 
northern Mexico : 

Archophileurus cribrosus (LeConte) 

Fig. 1, b 

Phileurus cribrosus LeConte, 1854, p. 80; Bates, 

1887, p. 338. 
Archophileurus cribrosus (LeConte) Casey, 1915, 
p. 264; Arrow, 1937, p. 87; Cazier, 1939, p. 170. 

Judged from collected specimens apparently 
the center of distribution of this species is in 



northern Mexico (Durango, Coahuila, and 
Tamaulipas), with further distribution in the 
southwestern United States (Texas, Arizona, 
and New Mexico). Nothing is known regarding 
its habits. The slender vestigial wings are inter- 
esting, and owing to this flightless condition we 
might expect to find local races in such a wide- 
ranging form. 

Genus Phileurus Latreille 

Phileurus Latrielle, 1807, p. 103; Burmeister, 
1947, p. 148; Lacordaire, 1856, p. 456; Kolbe, 
1910, p. 336; Casey, 1915, p. 264; Arrow, 1937, 
p. 89; Cazier, 1939, p. 170. 

In his 1937 catalogue Arrow lists 27 species, 
at least 10 of which are known to me to be 
synonyms. The species range generally through- 
out the Americas and the West Indies. The 
larvae live in decaying wood. Because of the 
flattened dorsal surface and the black color, 
these Phileurus are often mistaken for passalid 
beetles, which they do indeed superficially re- 

Phileurus truncatus (Beauvois) 
Fig. 1, a 

Scarabaeus truncatus Palisot de Beauvois, 1807, 

p. 41. 
Phileurus truncatus (Beauvois) Casey, 1915, p. 

265; Bates, 1889, p. 340; Arrow, 1937, p. 90; 

Cazier, 1939, p. 170. 
Phileurus recurvatus Casey, 1915, p. 266. 

Recorded by Bates and Casey from Mexico 
and ranging also rather commonly throughout 
our southeastern United States. Has been re- 
corded as mistaking chimneys for hollow trees 
and thus falling into fireplaces. The large size 
and strong cephalic horns readily place the 

Phileurus illatus LeConte 
Fig. 1, c 

Phileurus illatus LeConte, 1854, p. 80; Casey, 
1915, p. 267; Ritcher, 1944, p. 47 (larvae). 

Phileurus vitulus LeConte, 1866, p. 80; Cazier, 
1939, p. 170. 

Phileurus phoenicis Casey, 1915, p. 267; Cazier, 
1939, p. 170. 

Phileurus puncticollis Casey, 1915, p. 268; Cazier, 
1939, p. 170. 

Goniophileurus femoratus (Burmeister) Kolbe, 
1910, p. 149, p. 344 (pars); Arrow, 1937, p. 
86, 90; Blackwelder, 1944, pp. 257-258; Ca- 
zier, 1939, p. 170. 

There has been a great deal of controversy 
about and incorrect citations for this species 

since Kolbe erected the genus Goniophileurus 
for femoratus Burmeister and placed illatus 
LeConte and vitulus LeConte as synonyms of 
it. Burmeister's types of femoratus were from 
French Guiana and this name (the species is 
unknown to me) should apply to that locality, 
and vitulus and illatus should be removed from 
the synonymy of that species. Kolbe's main 
character for the genus Goniophileurus was the 
2- or 3-toothed mandible, whereas my dissec- 
tions show without doubt that the mandibles 
of our U. S. species are quite simple. Thus in 
the catalogues of Arrow (1937) and of Black- 
welder (1944) following Arrow, vitulus and 
illatus are listed both as synonyms of Gonio- 
phileurus femoratus (Burmeister) and also as 
valid species of Phileurus; actually, they have 
nothing to do with femoratus, and vitulus is a 
synonym of our common illatus. 

Ritcher has studied larvae taken in the 
trunks of trees (Dasylirion) in Arizona. The 
species occurs fairly commonly in Arizona, very 
rarely in southern California and in northern 
Mexico and Lower California (Triunfo, July 7, 
Ross and Michelbacher). I have also seen a 
specimen some time ago, apparently of this 
species, taken from the La Brea tar pits in 
southern California, probably representing a 
specimen of the (?) Pleistocene period. 

Phileurus castaneus Haldeman 
Fig. 1, d 

Phileurus castaneus Haldeman, 1843, p. 304; 
Casey, 1915, p. 270; Arrow, 1937, p. 89; Ca- 
zier, 1939, p. 170; Ritcher, 1944. p. 42 (larvae). 

Phileurus valgus Olivier (nee Linnaeus), 1789, 
p. 43; Arrow, 1937, p. 89. 

Phileurus texensis Casey, 1915, p. 268; Cazier, 
1939, p. 170. 

Phileurus sulcifer Casey, 1915, p. 269; Cazier 
1939, p. 170. 

Phileurus fioridanus Casey, 1915, p. 270; Cazier, 
1939, p. 170. 

Phileurus carolinae Casey, 1915, p. 269; Cazier, 
1939, p. 170. 

Arrow in his 1937 catalogue lists four varie- 
ties of this species from the West Indies and 
South America. It is a fairly common species in 
the United States, ranging from Virginia 
through the Southern States and Florida and 
Texas into Mexico, and supposedly also Gua- 
temala. Ritcher has reared the larva from a 
specimen collected in a cavity of a dead Bass- 
wood tree. 

May 15, 1948 delacour: note on aethopyga saturata (blyth) 



Arrow, G. J. Trans. Amer. Ent. Soc. 1909: 


. Coleop. Catalogus, pars 156: 89. 1937. 

. Trans. Ent. Soc. London (A) 86: 38. 

Bates, H. W. Biologia Centrali- Americana, 

Coleoptera 2(2): 338. 1889. 
Beauvois, A. M. J. Palisot de. Insectes 

recueillis en Afrique et en Amerique: 41. 

Blackwelder, R. E. U. S. Nat. Mus. Bull. 

185, part 2: 257-258. 1944. 
Bruch, Carlos. Rev. Mus. La Plata 4(2): 

73. 1910. 
Burmeister, H. Handbuch der Entomologie 

5: 148. 1847. 
Cartwright, 0. L. Ann. Ent. Soc. Amer. 

37(1): 36. 1944. 
Casey, T. L. Memoirs on the Coleoptera 6: 

178. 1915; 11:334. 1924. 
Cazier, M. A. Bull. Southern California 

Acad. Sci. 38(3): 170. 1939. 
Chevrolat, L. A. A. In Guerin, Mag. Zool., 

Coleopteres du Mexique, 13: 33. 1843. 
DeGeer, Carl. Memoires pour servir a Vhis- 

toire des insectes 4: 322. 1774. 
Duges, Eugene. Ann. Soc. Ent. Belgique 31 : 

137. 1887. 
Haldeman, S. S. Proc. Acad. Nat. Sci. Phila- 
delphia 1 : 304. 1843. 
Hamilton, J. Can. Ent. 18: 112. 1886. 
Horn, G. H. Trans. Amer. Ent. Soc. 2: 78. 

1870; 10: 122. 1882. 

Johnson, R. Connecticut Agr. Exp. Stat. 

Bull. 461: 79-86. 1942. 
Kirby, W. Trans. Linn. Soc. London 14: 

566-568. 1825. 
Kolbe, T. Ann. Soc. Ent. Belgique 54: 334. 

Lacordaire, J. T. Genera des coleopteres 3: 

456. 1856. 
Latreille, P. A. Genera crustaceorum et in- 

sectorum 2: 103. 1807. 
LeConte, J. L. Journ. Acad. Nat. Sci. Phila- 
delphia (2), 1:88. 1848. 
. Proc. Acad. Nat. Sci. Philadelphia 6: 

80. 1854; 8: 21. 1856; 13: 336. 1861. 
LeConte, J. L., and Horn, G. Classification 

of the Coleoptera of North America 259. 

Linnaeus, C. Amoen. Academy 6: 391. 1763. 
•. Sy sterna naturae (ed. 12) 1(2): 549. 

Manee, A. H. Ent. News 26: 266. 1915. 
Melsheimer, F. E. V. Proc. Acad. Nat. 

Sci. Philadelphia 2: 138. 1856. 
Olivier, A. G. Entomologie ou histoire na- 

turelle des insectes 1 (pars 5) : 43. 1789. 
Ritcher, P. O. Kentucky Agr. Exp. Stat. 

Bull. No. 467: 43. 1944. 
Say, Thomas. Proc. Acad. Nat. Sci. Phila- 
delphia 3: 209. 1823. 
Saylor, L. W. Journ. Washington Acad. Sci. 

35(12): 378-386. 1945; 36(1): 16-22. 

1946; 36(2): 41-46. 1946. 
Sims, Robert. U.S. Dept. Agr. Circ. 334. 1934. 
Sternberg, R. Stett. Ent. Zeitung 71: 26. 

1910 (1909). 

ORNITHOLOGY. — Note on the races of the black-throated sunbird, Aethopyga 
saturata (Hodgson). 1 J. Delacour, American Museum of Natural History. 
(Communicated by Herbert Friedmann.) 

While I was in Europe in the summer of 
1947, 1 made a complete examination of the 
specimens of Aethopyga saturata preserved 
in the Museums of Paris and London at the 
instance of H. G. Deignan, who has re- 
cently revised the races of the species, using 
the material available in the United States. 
I found that the races stand as he has indi- 
cated in his recent paper (This Journal 38 : 
21-23. 1948) with one addition. Also new 
indications are supplied by the specimens in 
the Paris and London collections, many of 
which have been collected by me in Indo- 
china. The metallic blue, or dull black, 
coloration of the middle of the throat of the 
males is an important characteristic, but it 
is not quite stable, and specimens varying 

1 Received January 30, 1948. 

in that way may be found apparently in 
several populations. I have listed the fol- 
lowing ratios of metallic and dull throats in 
two subspecies from the specimens de- 
posited in Paris and London. 

Subspecies and locality 

Metallic throat 

Dull throat 

sanguinipectus : 

Karen ni 


(including type) 





Southern Shan 



Haut Laos 






Even in specimens with a completely 
metallic blue throat the center is always 
somewhat duller than the sides, so that in 
some cases it is a question of degree and 
there is a gradation between the two ex- 



tremes. Also there is a good deal of variation 
in the intensity and extent of the yellow of 
the underparts among specimens from 
Tongking and Haut Laos. Two males from 
Dakto, Central Annam, are similar to speci- 
mens from the Boloven Plateau and should 
be referred to ochra. 

Mr. Deignan was unable to examine ma- 
terial from Bokor, southwestern Cambodia. 
This population is isolated on the Chaines 
des Elephants et des Cardamomes and has 
special characteristics. I propose to call it: 

Aethopyga saturata cambodiana, n. subsp. 

Type.— Brit. Mus. Nat. Hist. no. 1578, 
adult male, collected at Bokor, Cambodia, on 
December 12, 1927, by J. Delacour (original 
number 793). 

Diagnosis. — Nearest to ochra Deignan (Bas 
Laos and Central Annam), differing in having 
the middle of throat usually metallic blue, not 
dull black, the mantle of a darker maroon red 
color and the abdomen grayer. Resembles 
sanguinipectus Walden (South Burma), but 
darker red on the mantle. In its dark mantle 
and metallic throat it approaches johnsi Robin- 
son and Kloss (South Annam), which, how- 
ever, is very distinct from all other races in the 
almost plain red color of the breast. Iris dark 
brown; bill black; legs blackish brown. Female 
similar to peter si, very yellow underneath. 

Range. — The mountains of southwestern 
Cambodia and probably the border of Siam. 

Specimens examined. — 8 males, 3 females. 

Remarks. — One of the males (Paris) has the 
center of throat dull black. 

ORNITHOLOGY. — Some races of the babbling thrush, Malacocincla abbotti Blyth? 
H. G. Deignan, U. S. National Museum. 

Despite the fact that this common bab- 
bler is generally admitted to show normal 
subspecific variation in the Malaysian Sub- 
region, ornithological writers have con- 
sistently held that the nominate race 
ranges, without the least geographical 
change, from the eastern Himalayan foot- 
hills to Malaya and Indochine. Inasmuch as 
the rich material before me shows unde- 
niable subspeciation, it must be supposed 
that lack of specimens from Arakan, the 
type locality of the species, has inhibited its 
proper study in the past. 

It may be said at once that I have not 
myself seen a single topotype of Malaco- 
cincla abbotti and that all remarks to follow 
are based upon the premise that specimens 
from southwestern Siam and northern 
Tenasserim represent the Arakanese form 
— an assumption supported by careful com- 
parison of these birds with the original de- 
scription of Blyth (Journ. Asiat. Soc. Bengal 
14, pt, 2: 601. Aug. 1845). 

In my diagnoses of new subspecies, only 
fresh-plumaged adult examples have been 
employed, and "foxing" has been taken into 
account by comparison of birds of approxi- 
mately the same date of collection. After 
these precautions, I still find it necessary 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received January 
16, 1948. 

to name three races from Siam alone. Their 
descriptions follow. 

1 . Malacocincla abbotti ruf escentior, n. subsp. 

Type.— U. S. N. M. no. 330572, adult female, 
collected at Ban Tha Lo, southwest of Surat 
Thani or Ban Don {ca. lat. 9°05' N., long. 
99°15' E.), peninsular Siam, on September 20, 
1931, bjr Hugh M. Smith (original number 

Diagnosis. — Separable in series from M. a. 
abbotti (as exemplified by birds from south- 
western Siam and northern Tenasserim) by 
having the upperparts slightly more rufescent, 
and especially by having the underparts (ex- 
cept the white throat and abdomen) more 
strongly washed with a much brighter ferrugi- 

Range. — Peninsular Siam (except Pattani 
Province) and southern Tenasserim. 

Remarks. — M . a. rufescentior is distinguish- 
able from olivacea of Pattani Province and 
Malaya by the same characters as separate it 
from abbotti. The material before me does not 
show any very obvious difference between 
abbotti and olivacea, although the latter seems 
to have the upperparts the least bit darker in 
tone; larger series would probably show this 

Twenty-four winter-taken adults of rufes- 
centior have been examined. 

May 15, 1948 



2. Malacocincla abbotti obscurior, n. subsp. 

Type.— 15. S. N. M. no. 333912, adult male, 
collected at Khao Sa Bap (lat. 12°35' N., long. 
102°15 / E.), Chanthaburi Province, southeast- 
ern Siam, on October 25, 1933, by Hugh M. 
Smith (original number 6545). 

Diagnosis. — Separable in series from M. a. 
rufescentior by having the coloration of the 
crown, especially anteriorly, darker and more 
olivaceous; by having the remaining upper- 
parts equally rufescent, but decidedly deeper 
in tone; and by having the rufescent of the 
underparts (excepting the white throat and 
abdomen) slightly brighter and deeper. 

Range. — Southeastern Siam. 

Remarks. — Twenty-five winter-taken adults 
of obscurior have been examined. 

3. Malacocincla abbotti williamsoni, n. subsp. 

Type.—V. S. N. M. no. 324357, adult male, 
collected at Sathani Pak Chong, eastern Siam 

at lat. 14°40 / N., long. 101°25' E., on November 
16, 1929, by Hugh M. Smith (original number 

Diagnosis. — Like M, a. obscurior in the dark 
coloration of the crown but easily distinguish- 
able from it in series by having the remaining 
upperparts olivaceous brown, but slightly suf- 
fused with rufescent, and by having the under- 
parts (except the white throat and abdomen) 
more lightly washed with a paler ferruginous. 

From M . a. abbotti, which it resembles be- 
neath, williamsoni is separable by the deeper 
tone of the more olivaceous-brown upperparts 
and the darker coloration of the crown. 

Range. — Eastern Siam and Laos (Vientiane). 

Remarks. — This race is named in honor of Sir 
Walter J. F. Williamson, C.M.G., the well- 
known student of Siamese ornithology. 

Eleven winter-taken adults of williamsoni 
have been examined. 

ZOOLOGY. — Two new millipeds of Jamaica. 1 H. F. Loomis, Coconut Grove, Fla. 

Late in January and early in February, 
1937, Dr. E. A. Chapin, curator of insects, 
United States National Museum, collected 
insects and members of lower groups in 
Jamaica. The millipeds included in this 
collection were sent to me for identification, 
there being eight species of which two ap- 
pear to be undescribed, one representing a 
new generic type. These two new millipeds 
are here described and the previously known 
species in the collection listed. All speci- 
mens have been deposited in the National 

Glomeridesmus angulosus, n. sp. 

One male (type) and six other specimens in 
bottle labeled only "Sifting fern gully, Feb. 
2," but probably collected at Moneague, where 
other collecting was done the same day. 

Diagnosis. — This is the smallest West Indian 
species of the genus and has the posterior cor- 
ners of more of the caudal segments produced 
into acute angles than any other species. The 
last male legs also are distinctive. 

Description. — Length of largest specimen, a 
female, with 21 segments, 4 mm, width 1 mm; 

1 Received January 16, 1948. 

largest male, with 20 segments, 3 mm long. The 
generally dark color of living animals probably 
is almost entirely derived from the internal 
organs showing through the quite transparent 
and colorless body wall noticeable in preserved 

The pit behind each antenna is circular and 
not opened on any side, nor is the antennal 
socket opened behind or below although there 
is a depression below it as in the Haitian G. 
jenkinsi Loomis. 

From segment 12 or 13 to segment 19 in- 
clusive the posterior corners are increasingly 
produced into acute angles as shown in Fig. 1. 

Basal joint of the legs with posterior margin 
minutely serrate. Pleurae with about three 
transverse ridges in front, the back margin 
smooth but with 6 to 8 minute, short, pro- 
jecting setae; inner posterior corner acute. 
Penultimate legs of male with basal joints di- 
rected outward, the three terminal ones bent 
caudad. Last male legs with only the two ter- 
minal joints projecting beyond the penultimate 
legs, modified as shown in Fig. 2. 

Siphonophora robusta Chamberlin 

A female, apparently of this species from 
Moneague, station 370, February 2. 



Rhinocricus sabulosus Pocock 

A female from Moneague, station 370, 
February 2, and several specimens from "under 
dung" at Mocho, February 16. 

Rhinocricus solitarius Pocock 

A male collected with R. sabulosus above. 

Rhinocricus sp. 

A young specimen from near White Horses, 
station 386, February 6. 

Spirostrophus naresi (Pocock) 

Numerous specimens from Bath St. Thomas, 
February 6 and 8. 

Figs. 1-2. — Glomeridesmus angulosus, n. sp.: 1, Segments 16 to 20, lateral view; 2, two apical joints 
of last male leg on left side, ventral view over penultimate leg. 

Figs. 3-6. — Xaymacia granulata, n. sp.: 3, Antenna and part of opposite socket; 4, head and first 
penultimate three segments, the nonsetiferous tubercles not shown; 5, segment 9 of male showing 
typical dorsal sculpture; 6, gonopods, ventral view. 

May 15, 1948 



Xaymacia, n. gen. 

Genotype. — Xaymacia granulata, n. sp. 

Diagnosis. — From the shape of the gonopods 
it does not appear that this genus has any close 
relatives in the known chelodesmid fauna of 
the West Indies or the mainland surrounding 
the Caribbean area. The ornamentation of the 
dorsum is not duplicated in other members of 
the family in the region. 

Description. — Body of the size and propor- 
tions of the common Orthomorpha coarctata 
(Saussure) with which species specimens were 
collected in several localities. Males more 
slender and with the dorsum natter than fe- 
males. Dorsum thickly granulate in addition 
to three transverse series of slightly larger 
setose tubercles on segments 1 to 19, inclusive. 

Head large, as wide as segment 1; a strong 
sulcus on the vertex; antennae separated by 
little more than the diameter of one of the 
sockets, geniculate at joint 4; joints 5 and 6 
with a group of sensory hairs on the outer side 
near apex. 

Produced posterior corners of lateral keels, 
from segment 2 to 18 inclusive, subequal in 
size. Pore formula normal, the pores opening 
outward from the margin of the carinae. Sterna 
sparsely hispid. 

Gonopods with the apical half of the poste- 
rior division slender, pointed, and curving be- 
hind and partly obscured by the anterior divi- 
sion which is biramose and with its apical half 
in a sigmoid curve. 

The generic name is in reference to the old 
name "Xaymaca" from which the modern 
name of Jamaica was derived. 

Xaymacia granulata, n. sp. 

From January 28 to February 8, 1937, 
numerous specimens were collected at the fol- 
lowing localities: Caymanas, along Rio Cobre, 
(male type); Annotto Bay; Half Way Tree; 
Hope Gardens; Bath St. Thomas. 

Description. — Length 16 to 18 mm; body 
parallel-sided from segment 1 to 16; males 
definitely more slender than females and the 
dorsum flatter, nearly horizontal; general size 
and color very similar to Orthomorpha coarctata 

Living color dark brown except for the cor- 
ners of segment 1 and the lateral carinae of 
succeeding segments which are light yellow, 

the color being restricted to the outer margin 
of the carina at the front of each segment but 
broadening to include the entire posterior cor- 
ner; last segment wholly brown; sterna, legs, 
preanal scale, and the anal valves colorless. 

Head almost as wide as remainder of body; 
strongly and evenly inflated, subglobose, with 
a very definite sulcus extending across the 
vertex to between the antennae; the vertex 
shining, glabrous behind, sparsely and finely 
hispid in front, the remainder of the head much 
more densely hispid with erect hairs varying in 
length from very short to others several times 
as long. Antennae close together near the front 
of the head, separated by little more than the 
diameter of one socket, shaped as shown in 
Fig. 3; joints 5 and 6 each with a small area of 
sensory hairs near apex on the outer side. 

First segment semicircular, strongly convex, 
with the posterior corners depressed, thin, 
horizontal, rather acute but not produced 
backward; surface densely scattered with 
small vesiclelike granules as high as broad, and 
three transverse rows of slightly larger setifer- 
ous granules, 12 of which are along the anterior 
margin, 10 in the median row and 8 to 10 some- 
what in advance of the posterior margin; a 
single seta projects outward from the margin 
just in advance of the posterior corner. 

Ensuing segments with granules and trans- 
verse rows of setiferous tubercles similar to 
those of segment 1 ; a pronounced transverse 
sulcus crosses the middle of each segment and 
the lateral carinae have one or two setae pro- 
jecting outward from the outer margin (Fig. 
4). Second segment with the outer margin of 
the keels slightly longer than on ensuing ones, 
the posterior corners produced backward in 
the same degree which remains uniform to 
segment 18, corners of segment 19 reduced to 
half size; from segment 5 backward the outer 
margin of the keels thickened and containing 
an elongate impressed area opening outward, 
this being much broader in the poriferous keels 
(Fig. 5). Pore formula normal. 

Last segment short, conical, abruptly nar- 
rower at apex, the dorsal surface lacking gran- 
ules except those bearing the setae, there being 
an anterior row of six of these and a posterior, 
subapical, row of four, the outermost actually 
being on the lateral surface. 

Preanal scale large, triangular, the posterior 
margin of segment 19 just in front of it with 6 to 



10 marginal setae. Anal valves moderately in- 
flated, the margins thinly elevated. 

Sterna sparsely hispid with long erect hairs. 
Sterna of fourth male legs with two rather 
large conical tubercles, other sterna and legs 
normal. Gonopods as shown in Fig. 6. 

Orthomorpha coarctata (Saussure) 

Numerous specimens collected at Annotto 
Bay, January 30; Half Way Tree, January 28 
and 31; Caymanas, on sandy beach along Rio 
Cobre, February 3. 


416th meeting of board of managers 

The 416th meeting of the Board of Man- 
agers, held in the Cosmos Club, March 15, 
1948, was called to order at 8 p.m. by the 
President, Dr. F. D. Rossini. Others present 
were: H. S. Rappleye, W. L. Schmitt, W. W. 
Diehl, F. M. Defandorf, C. F. W. Muese- 
beck, R. Bamford, W. A. Dayton, F. B. 
Silsbee, M. A. Mason, A. 0. Foster, L. A. 
Rogers, C. L. Garner, C. L. Gazin, and, by 
invitation, H. E. McComb, R. J. Seeger, and 
L. V. Judson. 

The President announced the appointment 
of a Committee on Science Legislation: J. E. 
Graf, Chairman, A. T. McPherson, W. W. 

It was reported that the Executive Commit- 
tee had agreed to accept an invitation to join 
with the Library of Congress, the American 
Council of Learned Societies, and the Founda- 
tion fcr Integrated Education in cosponsoring 
a memorial meeting in honor of the late Alfred 
North Whitehead at the Library of Congress on 
Sunday, March 21, 1948. 

The Executive Committee recommended to 
the Board that an allotment of $20 be made 
to the Membership Committee to cover ex- 
penses of office, including the preparation of 
mimeographed summaries of the new-member 
qualifications for presentation to the Board. 

The Chairman of the Committee on Meet- 
ings, Dr. R. J. Seeger, announced that the 
March meeting would be given over to the 
Academy Aw T ard winners for 1947. 

The secretary read the following report sub- 
mitted by a committee appointed to consider 
the creation of an office of President-Elect, 
increase in the permitted number of members, 
and the addition of two standing committees: 

The Committee met on 25 February 1948 in 
the office of Dr. Gazin at the National Museum to 
consider the questions referred to it by the Board 
of Managers, namely, the questions of creating 
the office of "President-Elect" of the Washington 
Academy of Sciences, of increasing the permitted 
number of members, and of adding the Committee 
on Awards for Scientific Achievement and the 
Committee on Grants-in-Aid for Research to the 
standing committees of the Board of Managers. 
The Committee recommends the creation of the 
office of President-Elect to promote continuity of 
policies and objectives of the Academy by ac- 
quainting the income president with the current 
business and the administrative routine of the 

The Committee regards with favor the proposal 
to raise the permitted number of members of the 
Academy since this would make possible an ex- 
pansion of the Academy more nearly in propor- 
tion to the growth of Washington as a center of 
science, and would at the same time provide addi- 
tional income for the Academy. However, the 
Committee feels that the increase in number 
should be a modest one. A large increase would 
create so many vacancies that there might be a 
danger of lowering the standards of admission of 
the Academy. The Committee suggests that the 
permitted number of active members be raised 
from 650 to 700 and the number of resident active 
members from 500 to 550. The Committee feels 
that such action would provide ample room for 
suitable candidates for several years to come, 
especially since there are at present about 20 
vacancies in the Academy. 

The Committee recommends increasing the 
number of standing committees of the Board of 
Managers from four to six to include the Commit- 
tee on Awards for Scientific Achievement and the 
Committee on Grants-in-Aid for Research. Both 
of these committees have been standing commit- 
tees in effect for the past several years. 

[There followed a list of suggested changes in 
the Bylaws and Standing Rules to carry out these 

The Board accepted the report and in- 
structed the Secretary to submit to a vote of 
the membership the recommended changes in 

May 15, 1948 

proceedings: the academy 


the Bylaws, with the emendation that the 
increase in the number of members permitted in 
the Academy be changed from 50 to 100. 
Changes in the Standing Rules were accepted 
with one amendment. These changes are re- 
quired to be presented at the next meeting of 
the Board for final approval. 

Upon further discussion of the composition 
and duties of the various standing committees, 
it was voted that the present Special Commit- 
tee that recommended the changes in the By- 
laws look into the matter of the tenure of 
membership in the standing committees to 
consider the suggestion that some arrangement 
be made to permit a rotation or carry-over of a 
certain number of each committee to the suc- 
ceeding year, in order to facilitate functioning 
of the committee and continuity in plans. 

The Secretary reported the request by Dr. 
0. E. Meinzer that H. Freeborn Johnston 
be reinstated as a member of the Academy. Mr. 
Johnston resigned in 1939 as a result of ill 
health. The Board voted to reinstate him to 

The Secretary reported the death of Dr. 
Nelson Horatio Darton, formerly with the 
Geological Survey, an original member of the 
Academy, on February 28, 1948; and of Dr. 
William Ralph Maxon, formerly curator of 
the U. S. National Herbarium, Smithsonian 
Institution, on February 25, 1948. 

The Board approved the request of Dr. 
Oscar Riddle to be placed on the retired list, 
effective December 31, 1946. 

The Senior Editor, Dr. J. I. Hoffman, re- 
ported that he had read the monograph re- 
ferred to the Board of Editors for comment 
and had found it to be in good order and 
worthy of publication. The monograph, The 
Parasitic Birds of Africa, by Dr. Herbert 
Friedmann, was then referred to the new 
Committee on Monographs for its recommen- 
dations. Dr. Hoffman then brought up the 
question of the proposed index to the first 40 
volumes of the Journal, and after some dis- 
cussion the President was authorized to ap- 
point a committee to consider the index and 
make recommendations on its publication. 

The Board accepted the nomination of Dr. 
T. Dale Stewart as a Vice-President of the 
Academy representing the Anthropological So- 
ciety, replacing Dr. W. N. Fenton who with- 
drew from this office because of his election to 

the Board of Managers. 

Item 9 of the recommendations of the Com- 
mittee to consider "various matters pertaining 
to the Journal and its improvement," carried 
over from the unfinished business of the previ- 
ous meeting of the Board, was again discussed. 
The President was authorized to appoint the 
committee recommended, i.e., to study the 
functions of the Academy and to formulate a 
program that will integrate these functions, in- 
cluding the Journal. The Board requested 
that this committee make its report by Janu- 
ary 1949. 

The meeting was adjourned at 10:15 p.m. 
C. L. Gazin, Secretary 


There follows a list of persons elected to 
membership in the Academy, by vote of its 
Board of Managers, since January 13, 1947, 
who have since qualified as members in accord- 
ance with the bylaws. (See also previous list 
in January 15, 1948, issue of the Journal.) The 
bases for election are stated with the names of 
the new members. 


Elected January 13, 194-7 

Joseph P. E. Morrison, zoologist, U. S. 
National Museum, in recognition of his scien- 
tific attainments in the field of malacology, es- 
pecially for his studies on the taxonomy, 
anatomy, and biology of the fresh-water mol- 

Elected October 6, 1947 

Earle K. Plyler, physicist, National Bu- 
reau of Standards, in recognition of his work on 
the structure of molecules from infrared spec- 
tra, chemical analysis by infrared absorption 
measurements, and properties of matter as ex- 
hibited by characteristic absorption spectra. 

Elected December 15, 1947 

Robert C. Cook, biologist, managing editor 
of the Journal of Heredity, in recognition of his 
services to biology, in particular his long-time 
distinguished editorship of the Journal of 

Ira A. Gould, Jr., chemist, University of 
Maryland, in recognition of his studies ou the 
chemistry of milk, especially of the chemical 



changes brought about by the application of 

Tirey F. Ford, chemist, Bureau of Dairy 
Industry, in recognition of his contributions to 
the ultracentrifugal measurement of micellar 
sizes and in particular his researches on the 
particle size of the proteins of milk. 

Harold H. Shepard, entomologist, U. S. 
Department of Agriculture, in recognition of 
his contributions to entomology, particularly 
the action of insecticides, the biology of stored 
products, insects, and the bibliography of the 

Paul R. Miller, plant pathologist, Bureau 
of Plant Industry, Soils, and Agricultural En- 
gineering, in recognition of his contributions to 
the science of plant pathology and in particular 
his researches on apple rusts and on plant- 
disease survey methods, including spore load 
studies, development of new techniques, epi- 
demiological studies, etc. 

W. Gardner Lynn, biologist, Catholic Uni- 
versity of America, in recognition of his studies 
in embryology and herpetology. 

Hugh T. O'Neill, botanist, Catholic Uni- 
versity of America, in recognition of his contri- 
bution to systematic botany, especially the 
Cyperaceae and Arctic plants. 

Lee Ling, plant pathologist, Food and Agri- 
cultural Division, United Nations Organiza- 
tion, in recognition of his contributions to the 
mycology and plant pathology of China. 

George D. Rock, physicist, Catholic Uni- 
versity of America, in recognition of his contri- 
butions to ultrasonics. 

Francis E. Fox, physicist, Catholic Uni- 
versity of America, in recognition of his work in 
ultrasonics, particularly on the absorption of 
ultrasonic waves in liquids. 

Joseph S. Caldwell, physiologist, U. S. De- 
partment of Agriculture, in recognition of his 
studies in the physiology of fruit and vegetable 

Frank L. Campbell, entomologist, editor of 
the Scientific Monthly, in recognition of his 
researches in entomology, in particular the 
physiology of insects in relation to toxicology. 

Richard S. Dill, engineer, Heating and Air 
Conditioning Section, National Bureau of 
Standards, in recognition of his outstanding re- 
search in the fields of heating, insulation, and 
air-conditioning of structures. 

John K. Taylor, chemist, National Bureau 
of Standards, in recognition of his work on 
electrode potentials, refractive index, and 

J. Brookes Knight, paleontologist, U. S. 
National Museum, in recognition of his contri- 
butions to the knowledge of Paleozoic Gastro- 

Doris M. Cochran, zoologist, U. S. Na- 
tional Museum, in recognition of her scientific 
work in taxonomic herpetology, especially for 
the Herpetology of Hispaniola, U. S. National 
Museum Bulletin 177, 1941. 

Elected January 12, 1948 

John C. Ewers, ethnologist, U. S. National 
Museum, in recognition of his research on the 
material culture and arts and crafts of the 
Plains Indians in historic times. 

Walter S. Diehl, engineer, Bureau of 
Aeronautics, Department of the Navy, in 
recognition of his contributions to aerody- 
namics and to aeronautics. 

Ashley B. Gurney, entomologist, Bureau 
of Entomology and Plant Quarantine, in recog- 
nition of his contributions to the taxonomy of 
the Orthoptera, Corrodentia, Neuroptera, and 

Harry A. Borthwick, botanist, Bureau of 
Plant Industry, Soils, and Agricultural Engi- 
neering, in recognition of his work in plant 
physiology and in particular his researches on 
photoperiodism in relation to plant morphology. 

Marion W. Parker, botanist, Bureau of 
Plant Industry, Soils, and Agricultural Engi- 
neering, in recognition of his work in plant 
physiology and especially on the relation of 
light and temperature to plant growth. 

Philip Brierley, plant pathologist, Bureau 
of Plant Industry, Soils, and Agricultural Engi- 
neering, in recognition of his contributions to 
plant pathology, particularly in clearing up the 
nature and complexity of the virus diseases of 
Lilium longiflorum. 

Wilbur D. McClellan, plant pathologist, 
Bureau of Plant Industry, Soils, and Agricul- 
tural Engineering, in recognition of his work on 
use of fungicides and interaction of nutritives 
and fungicides in the pathogenicity of certain 

Paul C. Marth, botanist, Bureau of Plant 
Industry, Soils, and Agricultural Engineering, 

May 15, 1948 



in recognition of his work in plant physiology, 
in particular his pioneering research on the ef- 
fects of growth regulators on plants and fruits. 
Floyd F. Smith, entomologist, Bureau of 
Entomology and Plant Quarantine, in recog- 
nition of his investigations of insects as vectors 
of viruses and the application of aerosols to in- 
sect eaemies of ornamental plants. 


Elected January 13, 194.7 

Agesilau A. Bitancourt, biologist, Insti- 
tuto Biologico Sao Paulo, Brazil, in recognition 
of his contributions to tropical researches in 
plant pathology, particularly in citrus diseases. 

Elected December 15, 1947 

Julian H. Miller, plant pathologist, Uni- 
versity of Georgia, Athens, Ga., in recognition 
of his researches in mycology, in particular his 
contributions on the comparative morphology 
and taxonomy of the Sphaeriales. 

Ernest H. Volwiler, chemist, Abbott 
Laboratories, North Chicago, 111., in recognition 
of his contributions to organic chemistry, par- 
ticularly the synthesis of organic medicinals. 


The Anthropological Society of Washington 
at its annual meeting held on January 7, 1948, 
elected the following officers: President, Wil- 
liam N. Fenton; Vice-President, W. Monta- 
gue Cobb: Secretary, Marshall T. Newman: 
Treasurer, John C. Ewers: Councilors to the 
Board of Managers, E. Wyllys Andrews IV, 
Stella L. Deignan, George M. Foster, 
William H. Gilbert, Jr., Gordon R. Willey; 
Representative to the Washington Academy 
of Sciences, T. D. Stewart. 

A report of the membership and activities 
of the Society since the last annual meeting 

Life members, 1; Active members, 62; Asso- 
ciate members, 20; Total, 83. This represents 
a decrease of six since last year. 

The members elected during the year were: 
Active members: Dr. E. W. Andrews, Comdr. 
Sydney Connor, J. R. Caldwell, Dr. N. 
Datta-Majumder, Dr. Philip Drucker, 
Miss E. C. Davis, Dr. C. E. Hutchinson, 
Dr. D. W. Lockard, Dr. W. 0. Negherbon, 

Dr. Philleo Nash, Dr. M. H. Watkins, 
Associate member: Dr. George M. Foster. 

The Society records its deep sense of loss at 
the death of Dr. Beatrice Bickel, member 
since 1933. 

Except for the joint meeting with the Wash- 
ington Academy of Sciences at the Cosmos 
Club on January 16, 1947, all regular meetings 
were held at the U. S. National Museum. The 
Program Committee for the year comprised 
Dr. Marshall T. Newman, chairman, and 
Dr. Margaret Lantis. 

Titles of papers presented before the regular 
meetings of the Society were: 

January 16, 1947, 739th meeting, Waldo R. 
Wedel, Archeology and the Missouri River 
development program (slides; refreshments 
served by Washington Academy of Sciences). 

February 18, 1947, 740th meeting, R. Rug- 
gles Gates, Human ancestry from a geneticist's 
viewpoint (slides). 

March 18, 1947, 741st meeting, Preston 
Holder, The Motilon Indians, an untouched 
tropical forest group in northwestern South 
America. (Published in this Journal 37: 417- 
427, 1947). 

April 15, 1947, 742d meeting, W. Montague 
Cobb, The American Negro in the light of 
modern physical anthropology (slides). 

May 23, 1947, 743d meeting (joint meeting 
with the Medico-Chirurgical Society of the 
District of Columbia), Wilton M. Krogman, 
Anthropology and race relations. 

October 21, 1947, 744th meeting, Father 
Ernst Worms, The natives of northwestern 
Australia — a contemporary picture of their 
language and culture (sound color film). 

November 5, 1947, 745th meeting, George 
M. Foster, The People of Tzintzuntzan — a con- 
temporary study of social and economic adjust- 
ment in Mexico (color films). 

December 3, 1947, 746th meeting, Eugene 
C. Worman, Jr., The neolithic period in 
India — new evidence on early cultural move- 
ments in Asia. 

The Society voted to freeze the Perpetual 
Building fund at $2,000 and to subsequently 
add the interest to dues and other income for 
running expenses. In this way the scope of activ- 
ities can be enlarged. It was decided also to 
govern annual expenditures by a budget esti- 
mate made at the start of each year. 


The report of the Treasurer follows: 


A.S.W. dues collected $ 112.00 

Back dues (A. A. A.) collected to reimburse A.S.W. for carrying delinquents . . 5 .00 

Interest, Perpetual Building Association 57 .38 

Dividends, Washington Sanitary Improvement Co 33.60 

Dividends, Washington Sanitary Housing Co 12 .00 

Interest, U. S. Savings Bond 12.50 

Sale of Old Series Anthropologists 7 .60 

S 240.08 

A. A. A. dues paid for Secretary and one life member 10 .00 

Expenses, speakers 60 .00 

Dues, Inter-American Society of Anthropology and Geography 3 .00 

Printing and mailing notices 77 .03 

Incidental meeting expenses 24 .00 

Miscellaneous expenses, Secretary and Treasurer - 9.64 


Balance $ 56.41 


Funds invested in Perpetual Building Association (with interest to Dec. 31, 

1947) SI, 956. 50 

21 Shares Washington Sanitary Improvement Co. (par value S10 per share) 210.00 

2 Shares Washington Sanitary Housing Co. (par value S100 per share) 200 .00 

U. S. Savings Bond, Series G 500 .00 

Cash in bank 261 .65 

Total as of December 31, 1947 S3 , 128 . 15 

Total as of December 31, 1946 3,071 .74 

Increase S 56.41 

Less bills outstanding: 

To A. A. A. (Subscriptions to American Anthropologist): 

For 1 member, 3 years, 1948-50 S 15 .00 

For 1 member, 1 year, 1948 5 .00 


Net increase S 36.41 

Marshall T. Newman, Secretary. 

Officers of the Washington Academy of Sciences 

President Frederick D. Rossini, National Bureau of Standards 

Secretary C. Lewis Gazin, U. S. National Museum 

Treasurer Howard S. Rappleye, Coast and Geodetic Survey 

Archivist Nathan R. Smith, Plant Industry Station 

Custodian and Subscription Manager of Publications 

Harald A. Rehder, U. S. National Museum 

Vice-Presidents Representing the Affiliated Societies: 

Philosophical Society of Washington Walter Ramberg 

Anthropological Society of Washington T. Dale Stewart 

Biological Society of Washington John W. Aldrich 

Chemical Society of Washington Charles E. White 

Entomological Society of Washington C. F. W. Muesebeck 

National Geographic Society Alexander Wetmore 

Geological Society of Washington. .-. William W. Rubey 

Medical Society of the District of Columbia Frederick O. Coe 

Columbia Historical Society Gilbert Grosvenor 

Botanical Society of Washington . . . . Ronald Bamford 

Washington Section, Society of American Foresters William A. Dayton 

Washington Society of Engineers Clifford A. Betts 

Washington Section, American Institute of Electrical Engineers 

Francis B. Silsbee 

Washington Section, American Society of Mechanical Engineers 

Martin A. Mason 

Helminthological Society of Washington Attrel O. Foster 

Washington Branch, Society of American Bacteriologists Lore A. Rogers 

Washington Post, Society of American Military Engineers . Clement L. Garner 

Washington Section, Institute of Radio Engineers Herbert Grove Dorsey 

Washington Section, American Society of Civil Engineers Owen B. French 

Elected Members of the Board of Managers: 

To January 1949 Max A. McCall, Waldo L. Schmitt 

To January 1950 F. G. Brick wedde, William W. Diehl 

To January 1951 Francis M. Defaxdorf, William N. Fenton 

Board of Managers All the above officers plus the Senior Editor 

Board of Editors and Associate Editors [See front cover] 

Executive Committee Frederick D. Rossini (chairman), Walter Ramberg, 

Waldo L. Schmitt, Howard S. Rappleye, C. Lewis Gazin 

Committee on Membership 

Harold E. McComb (chairman), Lewis W. Butz, C. Wythe Cooke, William 

W. Diehl, Lloyd D. Felton, Regina Flannery, George G. Manov 

Committee on Meetings Raymond J. Seeger (chairman), 

Frank P. Cullinan, Fred L. Mohler, Francis O. Rice, Frank Thonb 

Committee on Monographs: 

To January 1949 * Lewis V. Judson (chairman), Edward A. Chapin 

To January 1950 Roland W. BroWxNT, Harald A. Rehder 

To January 1951 William N. Fenton, Emmett W. Price 

Committee on Awards for Scientific Achievement (Karl F. Herzfeld, general chairman): 

For the Biological Sciences 

C. F. W. Muesebeck (chairman), Harry S. Bernton, Chester W. Emmons, 
Elmer Higgins, Mario Mollari, Gotthold Steiner, L. Edwin Yocum 

For the Engineering Sciences 

Harry Diamond (chairman), Lloyd V. Berkner, Robert C. Duncan, 
Herbert N. Eaton, Arno C. Fieldner, Frank B. Schbetz, W. D. Sutcliffe 

For the Physical Sciences 

Karl F. Herzfeld (chairman), Nathan L. Drake, Lloyd D. Felton, 
Herbert Insley, William J. Rooney, Robert Simha, Michael X. Sullivan 

Committee on Grants-in-aid for Research 

. .F. H. H. Roberts, Jr. (chairman), Anna E. Jenkins, J. Leon Shereshevsky 

Representative on Council of A. A. A. S Frank Thone 

Committee of Auditors 

William G. Brombacher (chairman), Harold F. Stimson, Herbert L. Haller 

Committee of Tellers 

...John W. McBurney (chairman), Roger G. Bates, William A. Wildhack 



Physics. — General survey of certain results in the field of high pressure 

physics. Percy W. Bridgman 145 

Science and Freedom: Reflections of a physicist. Percy W. Bridg- 
man 156 

Chemistry. — A method for the determination of certain metals pres- 
ent in minor concentration in various substances N. Howell 
Furman, C. E. Bricker, and Bruce McDuffie 159 

Ethnology. — Self-torture in the Blood Indian sun dance. John C. 

Ewers 166 

Paleontology. — On two previously unreported selachians from the 

Upper Cretaceous of North America. David H. Dunkle 173 

Entomology. — Synoptic revision of the United States scarab beetles 
of the subfamily Dynastinae, No. 4: Tribes Oryctini (part), 
Dynastini, and Phileurini. Lawrence W. Saylor 176 

Ornithology. — Note on the races of the black-throated sunbird, 

Aethopyga saturata (Hodgson). J. Delacour 183 

Ornithology. — Some races of the babbling thrush, Malacocincla 

abbotti (Blyth). H. G. Deignan 184 

Zoology. — Two new millipeds of Jamaica. H. F. Loomis 185 

Proceedings: The Academy 188 

Proceedings: Anthropological Society 191 

This Journal is Indexed in the International Index to Periodicals 

Vol. 38 

June 15, 1948 

No. 6 




James I. Hoffman 



Alan Stone 


Frank C. Kracek 



Lawrence A. Wood 


J. P. E. Morrison 


Elbert L. Little, Jr. 


Richard E. Blackwelder 


James S. Williams 


Waldo R. Wedel 


Irl C. Schoonover 


published monthly 



450 Ahnaip St. 

at Menasha, Wisconsin 


Entered as second class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a speoial rate of postage provided for in the Act of February 28, 1925. 

Authorised January 21, 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author* 8 Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate) : 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 . 50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8.00 .60 

Vol. 23 to current vol. (12 numbers per vol.) 7.50 0.90 

• Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 




Vol. 38 

June 15, 1948 

No. 6 

PHYSICS. — Mass spectra of hydrocarbons. 1 
of Standards. 

Fred L. Mohler, National Bureau 

Mass spectra were first observed by J. J. 
Thompson by passing canal rays through 
transverse electric and magnetic fields. His 
discovery in 1913 (1) that there appeared 
to be two isotopes of neon led Aston (0) 
to much more elaborate experiments. For 
many years interest centered in discovery 
of new isotopes and precision atomic-weight 
measurements. Studies of molecular mass 
spectra were largely an American develop- 
ment. (3) The purpose was to study the 
mechanism of ionization of polyatomic mole- 
cules. Later Hippie (4) at Westinghouse 
and Washburn (5) of the Consolidated 
Engineering Corporation developed this 
type of mass spectrometer into an instru- 
ment suitable for chemical analysis of gas 
mixtures. The requirements for such an 
instrument are accurate automatic elec- 
trical recording and reliable reproducibility. 

An important feature intioduced by 
Bleakney was the ionization chamber in 
which the ions were produced. A beam of 
univelocity electrons traverses the ioniza- 
tion chamber and a magnetic field parallel 
to the beam keeps this beam centered while 
a small transverse electric field draws the 
ions produced out of the ionization cham- 
ber. The ions are then accelerated by a 
large electric field and bent by a magnetic 
field to give a mass spectrum. 

The gas to be ionized enters the ioniza- 
tion chamber in a jet at very low pressure 
(about 10~ 4 mm), and fast pumps maintain 
a pressure of about 10 -6 mm outside the 
ionization chamber. This insures ideally 
simple conditions. There is no appreciable 
accumulation of ionization products in the 

1 Address of the retiring President of the Philo- 
sophical Society of Washington, January 17, 
1948. Received March 17, 1948. 

ionization chamber and there are no colli- 
sions between ions and gas molecules. As 
the applied voltage of the electron beam is 
increased, molecule ions first appear at 
about 10 to 12 volts. Then with increasing 
voltage, ions of various dissociation prod- 
ucts are formed. Twenty-five or 30 volts 
are sufficient to produce almost every pos- 
sible dissociation product of a hydrocarbon 
and, in the range 50 to 100 volts, the rela- 
tive intensities of the molecule ion and the 
various dissociation products remain nearly 
independent of voltage. The resulting mass 
spectrum is a property of the molecule modi- 
fied only slightly by instrumental factors. 

In the Consolidated mass spectrometer 
the ions from the ionization chamber are 
accelerated by a variable electric field and 
bent through an arc of 180° by the field of 
a large magnet to reach a collecting elec- 
trode. After passing through the electric 
field V all the singly charged ions will have 
the same kinetic energy 1/2 mv 2 = eV and 
light ions will have a high velocity and 
heavy ions a small velocity. The magnetic 
field exerts a force proportional to the veloc- 
ity at right angles to the field and to the 
velocity. Ions of mass m and charge e 
will move on the arc of a circle of radius 
R defined by the relation m/e = CH 2 R 2 /V 
where C is a constant, H is the magnetic 
field, and V the ion accelerating voltage. 
If V is gradually changed from high to low 
values, ions of different masses will succes- 
sively reach the ion collector. The current 
reaching the collector is amplified and re- 
corded by galvanometers on a moving 
sheet while l/V changes at a uniform rate. 
Thus, the resulting record shows a series of 
peaks at the values of l/V or m/e corre- 
sponding to the molecular weights of the 
compound and its dissociation products. 


% $/■ 



Fig. 1 is part of the mass spectrum record 
obtained with a Consolidated mass spec- 
trometer of n butane, CH 3 • CH 2 • CH 2 • CH 3 . 
Four galvanometers with four ranges of 
sensitivity record the ion current. The 
upper trace records the galvanometer of 
highest sensitivity. Deflections of the other 
galvanometer traces are to be multiplied 
by 3, 10, and 30 to give deflections in terms 
of the upper trace. The parent peak, the 
undissociated molecule ion, is at mass 58 
and a peak at 59 comes from molecules 
containing one C 13 isotope. Then there are 
molecules which have lost from 1 to 10 
hydrogen atoms. It is a remarkable fact 
that in a single ionization process all ten 
hydrogen atoms can be removed to give 
Ci + . The most probable ionization process is 
losing CH 3 to give mass 43. Losing an ad- 
ditional CH 2 to give 29 is also probable. 

One finds indeed every possible dissocia- 
tion piocess consistent with the structural 
formula. Peaks at 25 J and 26 J come from 
doubly charged ions of mass 51 and 53. 

The small broad peaks are metastable ion 
transitions resulting from ions which dis- 
sociate after traversing the electric field. 
Hippie, Fox, and Condon (6) have studied 
and explained these metastable transition 

The use of mass spectra for chemical 
analysis involves the direct comparison 
of the mass spectrum of an unknown with 
the spectra of pure compounds and an im- 
portant project of the Mass Spectrometry 
Section is compiling mass spectra of pure 

To show characteristics of the mass 
spectra of hydrocarbons it is convenient 
to use a greatly simplified picture in which 
only peaks greater than 2 percent of the 
maximum peak are included. The maximum 
peak is uniformly given a height of 100. 
First are shown spectra of saturated hydro- 
carbons of formula C n H 2n+2 . Fig. 2 shows 
some of these. 

In methane the most probable process is 
ionization without dissociation. Ionization 




36 40 

Fig. 1. — Part of the mass spectrum record of n-butane as obtained 
by a Consolidated mass spectrometer. 


June 15, 1948 

Mohler: mass spectra of hydrocarbons 


with removal of one hydrogen atom is prob- 
able but removal of H+ is improbable and 
this is true in all hydrocarbons. In ethane 
the most probable ionization process in- 
volves removal of two hydrogen atoms. 
Production of CH 3 + is relatively improbable 
here and in all hydrocarbons except meth- 
ane. In propane and the two butanes loss 
of CH 3 is the most probable ionization 

It is a useful property of the structure of 
saturated hydrocarbons that the parent 
peak of each hydrocarbon is characteristic 
of that hydrocarbon and can not be pro- 
duced as a dissociation product of other 
compounds. The other mass peaks recur 
in spectra of heavier compounds. This is 
very convenient for the purpose of chemical 

There are two isomers of butane and the 
different structures give markedly different 
spectra. One difference between n-butane 

80 — 

I I I 




I I I I I I I I I I 



! IT I I 

Mil I' 

I I I I I I i i i 

CH 5 CH 3 

i r 

1 I 1 I I I I I 

CH 3 CH 2 CH S 





I I I I I I 

.jun 2 «.. jU „ 3 . 

I I I I I I 

CH 3 CHCH, _ 

Fig. 2. — Mass spectra of methane, ethane, 
propane, and two butanes. 

20 30 40 50 60 70 80 

1 1 I I 

I l""l I ' I 

— i — r - 1 — i i i 

CH^GHoGHoG HavH * 

I I I i I I J 


I i r"i i t"i i '1 

i i i i i 

CH 3 


Fig. 3. — Mass spectra of pentanes. 

and iso-butane is that one can not break 
iso-butane in half without first rearranging 
the hydrogen atoms. A peak at mass 29 
is found so this rearrangement must occur 
before dissociation. Such rearrangements 
are common in more complicated molecules. 
The parent peak is much smaller in iso- 
butane and it is a general rule that mole- 
cules with side chains have smaller parent 
peaks than the normal molecule. 

Fig. 3 shows spectra of the three isomers 
of pentane. In normal and iso-pentane losing 
CH2+CH3 is the most probable process. In 
neopentane losing CH 3 is most probable 
and the parent peak is only a few hundredths 
of a percent. Losing two CH3 radicals is 
quite improbable and the loss of CH3+CH4 
involving breaking of three bonds is quite 
probable. The peak at C2H 5 + is large al- 
though this requires a rearrangement of 
hydrogen atoms before dissociation, as in 
iso-butane. Here the spectrum gives little 
clue as to the structure of the molecule. 

There are five hexanes shown in Fig. 4 
and these give five very different mass 
spectra. Many of the features of these 
spectra can be described qualitatively by 
the statement that there is a tendency to 
dissociate at either side of each side chain. 
Normal hexane like normal pentane loses 
C 2 H & most readily. In 2-methyl-pentane, 
loss of one or three carbon groups is pre- 
ferred, while in 3-methyl-pentane, loss of 



10 tO 50 40 SO 60 70 80 90 

1 — I — i — i — I i I i — T 

I l "" l I 'I 

i r'i I ' | " 'l I 1 

i r 

i — i — i — i — i — i — i — r 


J r _ 

"i — n — i — rn — r 



» w "t wr, » 

I l ""l I 'I 


i i i i i i i r 


i i'Ii i rn i i"* i r 


I I I I I I I I I' I I 



Fig. 4. — Mass spectra of hexanes. 

C2H6 is again preferred. In 2,3-dimethyl- 
butane, loss of one and three are preferred 
while in 2,2-dimethyl-butane, one and two 
are preferred. In the last case, however, 
the most probable ionization process is 
breaking in half to give 43+ although this 
involves a rearrangement of hydrogen 
atoms. Peaks involving rearrangement are 
also noted at 43 in 3-methyl-pentane and 
at 29 in 2,3-dimethyl-butane. We know 
that these cases involve rearrangement be- 
cause the original structure does not yield 
such masses but obviously rearrangement 
may occur in other cases where it is indis- 
tinguishable from simple dissociation. 

Our measurements include the nine iso- 
mers of the heptanes and the 18 octanes. 
All these show T a great variety of mass spec- 
tra depending on the structure of the mole- 
cules. The tendency to dissociate on either 
side of a side chain is again found except 
in cases where rearrangements give ions 
which would not be expected on the basis 
of the molecular structure. 

Fig. 5 shows some unsaturated molecules 
with double or triple bonds. In ethylene the 
parent peak is the maximum peak, while in 

ethane it is about 25 percent of the maxi- 
mum. In acetylene the parent peak is by 
far the largest peak. In propylene the parent 
peak is not the largest, but the whole C 
three group of ions is much larger than in 
propane. Propylene always loses CH 3 rather 
than CH 2 and this is true of most of the 
unsaturated molecules. There are two iso- 
mers of C 3 H 4 , propadiene and methyl-acety- 
lene, and their mass spectia are almost 
exactly alike in spite of the difference in 
structure. It seems very probable that in 
this case first a molecule ion is formed and 
then the hydrogen atoms are redistributed 
to give identical ions from both isomers. 
Fig. 6 shows some unsaturated C four 
molecules. There are four butene isomers 
and they all give mass spectra similar to 


10 20 30 



1 1 1 1 1 


1 1 1 1 1 

CH2=CH 2 — 



... .1 

— " 

' | 1 1 1 1 


! till 








■1 ' 


1 1 1 1 



1 1 1 1 

CH 3 CH=CH 2 _ 



■1 1 



I 1 1 1 1 1 


1 1 1 1 




C H JiC = C H2 — 








1 1 r"i 1 f 1 


1 1 1 1 




CH 3 C=CH — 






Fig. 5. — Mass spectra of ethylene, acetylene, 
propylene, propadiene, and methyl acetylene. 

Fred L. Mohler, president of the Philosophical Society of Washington, 1947 

June 15, 1948 



the 1-butene spectrum shown in the figure. 
2-butene, CH 3 -CH = CH-CH 3 , has a cis 
and a trans form depending on whether the 
two central hydrogens are on the same side 
or opposite sides of the chain. The mass 
spectra are nearly identical, the most con- 
spicuous difference being that the mass 29 
peak is 15.5 percent of the maximum in the 
cis compound and 19.8 percent in the trans 
compound. These two molecules cannot 
give a 29+ ion without a rearrangement of 
hydrogen atoms. 

There are four isomers of C4H 6 and the 
mass spectra show differences of a rather 
unexpected nature. The most probable 
ionization process in CH 2 :CHCH:CH2 is 
loss of mass 15 which involves a double 
dissociation while in the other isomers 
which can lose 15 in a simple dissociation 
this transition is less probable. Also, the 
mass 28 peak is largest in 1,3-butadiene 
where production of 28+ involves a rear- 
rangement of hydrogen atoms. 

The unsaturated hydrocarbons show 
several characteristics as a class. The parent 
peaks are larger than for saturated com- 
pounds and larger for doubly unsaturated 
compounds than for mono-olefins. Ioniza- 
tion by loss of CH 2 is very improbable even 
where both terminal radicals are CH 2 . Both 
properties reflect the fact that the unsatu- 
rated bonds are stronger than the saturated 
bonds. Mass spectra of unsaturated mole- 
cules are also much less sensitive to struc- 
tural differences among different isomers 
than is the case with saturated molecules. 
This indicates that rearrangements of hy- 
drogen atoms in the molecule ion occur more 
readily in the unsaturated molecules. It 
may seem surprising that there are any 
rearrangements in saturated hydrocarbons 
and it will be of interest to study mass 
spectra with deuterium substituted at one 
position in the molecule and see where it 
appears in the mass spectrum. 

On the basis of momentum considerations 
it must be assumed that electron collisions 
always produce molecule ions and that 
these ions may be left in a highly excited 
state and subsequently dissociate spon- 
taneously to give the great variety of ions 
observed in the mass spectrum. I have re- 
ferred before to small wide peaks in the 



I i i — i — r~r 

i — I ! I II 

i i i i i i i 

CH,CH 2 CH:CH 2 _ 

i — I I I I I — r 

i — r 


~i — I I l 

CH 2 S C*CHCH 3 

i — i 'I i Pi — rn~ t 

i r 


T— T 



i i i i i 

CH 2 ?CHCH«CH 2 . 

m — I I I 

CH=CCH e CH 3 _ 

I I I I I 
CH 5 C = CCH 3 . 

Fig. 6. — Mass spectra of butene-1, 
butadienes, and butynes. 

mass spectrum that arise from ions which 
dissociate after they have traversed the 
electric field. Because of this phenomenon 
of delayed dissociation we can obtain direct 
experimental evidence as to some of the 
dissociation processes which occur. 

The apparent mass m a of an ion which 
dissociates immediately after traversing the 
electric field is 

m a = mf/nti 
where m* is the initial mass and m f the 
final mass of the ion (6). Because the ions 
dissociate over a range of positions the 
peaks are wide. The dissociating ions are 
presumably meta stable ions with a life of 
the order of 10 -6 seconds (7). As m f and ra t - 
are integers equal to or less then the molec- 
ular weight the numerical value of m a 
is sufficient to determine both m f and ra t -. An 
important aid in finding m f and m»- is a 
qualitative intensity rule that the mass 
peaks corresponding to m f and m* are always 
fairly large peaks in the mass spectrum. 



To return to Fig. 1, there are seven me- 
tastable transitions in the range of this 
record but some are very hard to see. The 
one at mass 32 comes from the parent ion 
of mass 58 losing mass 15, a CH3 radical. 
The one near mass 30 comes from 58 losing 
CH4. These two transitions only occur in 
n-butane. A large peak near 39 comes from 
43+ losing two hydrogen atoms and near 
25 is a peak from 29 losing two hydrogens. 
These last two recur in many hydrocarbons. 

Mrs. Bloom of our Section has made a 
careful compilation of mass spectra of 56 
hydrocarbons and has found 362 cases of 
metastable transition peaks and has identi- 
fied the transitions involved (£). The com- 
pounds include saturated hydrocarbons 
through C eights, mono-olefins through C 
fives, and four C 4 H 6 isomers. The data have 
been published in the American Petroleum 
Institute Catalogue of Mass Spectral Data 

The transitions frequently recur in dif- 
ferent hydrocarbons and 32 different tran- 
sitions are found. These involve loss of 
masses ranging from 2 to 44. Table 1 
summarizes the metastable transitions. 
Loss of mass 2 is by far the most common 
type of metastable transition and it appears 
in nearly all cases where there are two large 
peaks differing by two mass units. Losses 
of masses 1 or 3 are never observed. 

It is also significant that there are no 
metastable transitions involving loss of 
CH2 as this is a structural unit in most hy- 
drocarbons and large peaks often fall at 
intervals of 14. 

Loss of CH 3 is only observed from parent 
ions and is a comparatively infrequent 
phenomenon. CH4 is not a structural unit 
and loss of 16 involves a double dissociation. 
The loss from the ion 57+ recurs in 30 of 
the 40 saturated hydrocarbons. 

Loss of C 2 H 2 seems to be a unique prop- 
erty of the 55+ ion and recurs in most of 
the hydrocarbons. The loss of C 2 H 3 occurs 
only when the butadienes and butyne-2 
split in half. It is of interest that the un- 
symmetrical molecule CH 2 = C = CH — CH 3 
splits in half although this requires a re- 
arrangement before splitting. C2H4 in the 
form CH 2 • CH 2 or CH • CH 3 is a structural 
unit of many saturated molecules and the 

2 metastable transitions recur in many of 
them. In this case, the terminal CH 3 radical 
must first be removed from the initial ion. 

The loss of mass 29 comes from loss of 
the structural unit CH 2 CH 3 and includes 
the terminal radical as is shown by the fact 
that it comes from parent ions. Loss of 
mass 30 comes from losing two CH 3 radicals 
from the parent ion and necessarily involves 
a double dissociation. 

The mass C?H 6 can be CH 2 • CH 2 • CH 2 or 
can involve methyl side chains, but like 
C 2 H 4 it will never come from parent ions. 
Of course, it can only come from rather 
large molecules. Loss of mass 44 from the 
parent mass of 3 heptanes is perhaps doubt- 
ful. We are not certain that there is not an 
alternative explanation. It involves a double 
dissociation, removal of mass 43 and one 

This looks very complicated, and I have 
given it in detail to show that dissociation 
of ions does indeed involve many alterna- 
tive processes. If the initial metastable ion 
is the parent ion this means that the mole- 
cule ion traverses the electric field and then 
dissociates. When the initial ion is not the 
parent ion then the dissociation takes place 
in at least two stages. The initial metastable 

Table 1. — Summary of Metastable Transitions 

Mass lost 




8 transitions in all hydrocarbons except 



7 transitions from parent ions of 13 unsatu- 
rated and 5 saturated hydrocarbons. 



55 h -+39 + +16 in pentenes. 
57 + -*41 + +16 in 30 saturates. 
58+_^42+ -1-16 in n-butane. 


C 8 H, 

55 + ->29 + +26 in 41 hydrocarbons. 



54 + -»27 + +27 in 3 isomers of CtUt. 


C 2 H, 

71 + -43 + +28 in 21 saturates. 
85 + -*57 + +28 in 14 saturates. 



3 transitions from parent ions of pentenes, 
n-hexane, and 3 octanes. 



4 transitions from parent ions of 12 satu- 
rated hydrocarbons. 


C 3 H» 

85 + — *43 + +42 in 19 heptanes and octanes. 
99 + -»57 + +42 in 17 octanes. 



100 + ->56 + +44 in 3 heptanes. 

June 15, 1948 Larsen: new species of achaetogeron from Mexico 


ion is formed by dissociation within the 
ionization chamber, then it passes through 
the electric field and dissociates again. 
Sometimes large structural units are broken 
off in a single dissociation; in other cases at 
least two bonds must be broken in the de- 
layed dissociation as in losing two CH 3 radi- 
cals or CH 4 . 

The metastable transitions account for 
only a small fraction of all the dissociations. 
Hippie (7) has made quantitative estimates 
in the case of n-butane. The parent ion 
gives rise to two metastable transitions 
with loss of masses 15 and 16. The life of 
each excited state is about 2X10 -6 sec 
and initially at time zero not over 10 per- 
cent of the 58 + ions are in the excited state 
or states which give rise to these transitions. 
Now the 58+ ions account for only about 
4 percent of all the n-butane ions. Ninety- 
six percent of all the ions dissociate im- 
mediately or very quickly, about 3.6 per- 
cent become stable 58 + ions and 0.4- per cent 
fall into the metastable states. This is prob- 
ably a typical case as far as orders of magni- 
tude are concerned. 

What I have said has been largely de- 
scriptive and probably seems like little more 
than a catalog of empirical facts. However, 
this is largely a matter of viewpoint. As I 
pointed out in the introduction, these mass 
spectra are a molecular property and are 

somewhat analogous to intensities in molec- 
ular absorption spectra but in absorption 
spectra there is a theoretical basis for the 
interpretation of results. In spite of all the 
work on mass spectra we still lack this 
theoretical background. This is not only a 
challenging problem but it is of considerable 
practical importance. Chemical analysis 
now depends on comparing unknowns with 
pure samples of the various components. 
It would be very useful if we could deduce 
the structure of pure compounds of un- 
known structure. Beyond C sevens the 
possibilities become so many that empirical 
methods can never give the complete an- 


(1) J. J. Thompson in an address to the Royal 

Institution, 1913. 

(2) Aston. Mass spectra and isotopes. Lon- 

don, 1942. 

(3) Smyth. Rev. Mod. Phys. 3: 347. 1931. 

(4) Hipple. Journ. Applied Phys. 13: 551. 


(5) Washburn, Wiley, and Rock. Ind. 

Eng. Chem., Anal. Ed. 15: 541. 1943. 

(6) Hipple, Fox, and Condon. Phys. Rev. 

69: 347. 1946. 

(7) Hipple. Phys. Rev. 71: 594. 1947. 

(8) Bloom, Mohler, Lengel, and Wise. (In 

press Journ. Res. Nat. Bur. Standards). 

(9) American Petroleum Institute, Re- 

search Project 44. 

BOTANY.- — New species of Achaetogeron (Compositae) from Mexico. 1 Esther 
L. Larsen, Crown Point, Indiana. (Communicated by S. F. Blake.) 

The new species here described were 
discovered some years ago when the genus 
Achaetogeron was being studied together 
with several related genera of the tribe 
Astereae. The work was done at the Mis- 
souri Botanical Garden. The following ab- 
breviations are used for herbaria cited: 
M = Missouri Botanical Garden; NY = New 
York Botanical Garden; Phil = Academy of 
Natural Sciences, Philadelphia; US = U. S. 
National Herbarium. 

Achaetogeron filiformis Larsen, sp. nov. 

Annuus 15-45 cm altus; caules tenues juven- 
tate simplices maturitate diffuse ramosi foliosi 
patenti-hirsuti; folia obovata ad obovato- 

1 Received March 23, 1948. 

spathulata pinnatifida 0.5-4 cm longa, lobis 
obtusis apiculatis strigosis; folia suprema brac- 
teiformia oblonga integra; pedunculi filiformes 
1.5-2.5 cm longi infra capitulum dense patenti- 
hirsuti; capitula 0.7-1 cm diam.; phyllaria 2-3- 
seriata lineari-lanceolata acuminata glandulari- 
puberula basi hirsuta; radii albi vel caerulei 
supra discum arete circinnati; pappus incon- 
spicuus coroniformis laceratus, in floribus radii 
interdum etiam setis paucis donatus; achenia 
pubescentia pilis apice rectis vel obscure biden- 

Tamaulipas: Vicinity of Tampico, alt. 15 
m, March 10-April 19, 1910, E. Palmer 249 
(type, M, also US); vicinity of La Barra, 8 
km east of Tampico, at sea level, February 
1-8, 1910, E. Palmer 281 (M, US); en route 



from San Luis Potosi to Tampico, December 
1878 to February 1879, E. Palmer 1089 (US). 
Veracruz: Vicinity of Panuco, April 20-25, 
1910, E. Palmer 356 (US). 

The filiform peduncles, small heads, and 
circinate rays which scarcely exceed the disc 
distinguish this species. 

Achaetogeron fisheri Larsen, sp. now 

Herbaceus 20-45 cm altus ubique strigosus, 
ramis adscendentibus e caule decumbente sub- 
lignoso; folia caulina sessilia alte pinnatisecta 
1-3 cm longa lobis 3-7 lineari-lanceolatis 
obtusis, folia suprema saepius integra linearia; 
pedunculi 2-4 cm longi strigosi; capitula 2-2.5 
cm diam. (radiis inclusis); involucri 2-seriati 
0.5 cm alti phyllaria lineari-lanceolata acumi- 
nata glanduloso-puberula et strigosa, margine 
scariosa; pappus annularis laceratus incon- 
spicuus cum setis caducis interioribus; achenia 
basi callosa pubescentia; pilis bidentatis. 

State of Mexico: Amecameca, alt. 2,648 
m. July 29, 1924, Fisher (M, No. 914802 type); 
Amecameca, alt. 2,676 m, July 29, 1924, Fisher 
220 (US). Puebla; Teutla, October 1913. 
Salazar (US); Manzanilla, vicinity of Puebla, 
November 24, 1908, 10184 (US). 

The subpinnatifid leaves and a pappus crown 
so reduced as to be nearly lacking are charac- 

Achaetogeron sophiaefolius Larsen, sp. nov. 

Perennis 30-60 cm altus; caules plures e basi 
sublignoso ramosi striati patenti-hirsuti et 
subglandulari-puberuli ; folia elliptica ad ovata, 
pinnatisecta vel bipinnatisecta inferiora petio- 
lata usque ad 9 cm longa pilis longis multicel- 
lularibus appressis pubescentia et minute glan- 
dulari-puberula, segmentis ultimis obtusis, 
superiora sensim reducta sessilia, suprema 
bracteiformia; pedunculi 1.5-3 cm longi dense 
appresse pubescentes; capitula terminalia soli- 
taria 2-2.5 cm diam. (radiis inclusis) ; involucri 
1-1.2 cm diam. phyllaria 2-seriata 4 mm longa 
lineari-lanceolata acuminata glandulosa et parce 
hirsuta pilis multicellularibus, margine mem- 
branacea; pappus minutus coroniformis lace- 
ratus cum setis paucis caducis interioribus; 
achenia basi callosa pubescentia pilis apice 
rectis vel dentatis. 

Durango: City of Durango and vicinity. 
April-November 1896, E. Palmer 158 (M, No. 
123017 type, NY, US). 

A relatively coarse plant with conspicuously 
bipinnatisect leaves, which are pubescent with 
long appressed multicellular hairs. 

Achaetogeron pringlei Larsen, sp. nov. 

Herbaceus erectus ca. 65 cm altus ubique 
glandulari-puberulus et patenti-hirsutus; folia 
caulina sessilia alte pinnatisecta lobis ca. 5-9 
linearibus vel lineari-spathulatis obtusis utrim- 
que sparse villosa margine ciliata; folia in- 
florescentiae lineari-lanceolata integra vel inter- 
dum lobata sensim reducta, suprema ca. 1 cm 
longa; pedunculi 1.5-2.5 cm longi dense glandu- 
lari-puberuli pilis sparsis intermixtis; capitula 
1.5-2.5 cm diam.; involucri 2-seriati phyllaria 
lineari-lanceolata acuminata anguste mem- 
branaceo-marginata glandulari-puberula et 
sparse pilosa; radii numerosi albi saepe reflexi; 
flores disci numerosi; pappus minutus coroni- 
formis laceratus cum setis paucis interioribus 
caducis; achenia anguste obovata basi minute 
callosa pubescentia pilis apice rectis vel biden- 

Chihuahua: Cool slopes, Sierra Madre, 
October 10, 1888, Pringle 1625 (M, type); 
cool slopes, foothills of the Sierra Madre, 
October 11, 1887 Pringle 1272 (Phil). Puebla: 
Santa Maria de Zacatepec, vicinity of Puebla 
December 1908, Arsene (US). 

Distinguished by deeply pinnatisect, spar- 
ingly villous leaves and a relatively finer habit 
than A. sophiaefolius. 

Achaetogeron polycephalus Larsen, sp. nov. 

Herbaceus lm altus et ultra (basi inviso) 
ubique strigosus et minute glandulari-puber- 
ulus; caulis supra valde ramosus aetate glabra- 
tus; folia caulina majora oblanceolata sessilia 
prope apicem serrata usque ad 5 cm longa 1.5 
cm lata, minora valde reducta integra; pedun- 
culi 1-4 cm longi striati strigosi et minute 
glandulari-puberuli; capitula 2-3 cm diam. 
(radiis inclusis) ; involucri 2-seriati phyllaria 
lineari-lanceolata acuminata anguste mem- 
branaceo-marginata minute glandulari-puber- 
ula basi patentipilosa margine angusto mem- 
branaceo leviter lacerato; radii numerosi albi; 
pappus minutus annularis laceratus cum paucis 
setis interioribus; achenia basi callosa dense 
pubescentia pilis apice rectis vel bidentatis. 

Durango: San Ramon, April 21-May 18, 
1906, E. Palmer 52 (M, type). 

A profusely branching, strigose-pubescent 

June 15, 1948 



plant with serrate leaves and many scattered 

Achaetogeron corymbosus Larsen, sp. nov. 

Herbaceus 20-40 cm altus dense breviterque 
cinereo-hirsutus et minute glandulari-puberulus ; 
caules erecti ramosi ramis apice corymboso- 
ramosis; folia caulina oblanceolata sessilia 
supra serrata 5 cm longa 1.5 cm lata, superiora 
ad bracteas sensim reducta; folia infima non 
visa; pedunculi 0.5 — 1 cm longi; capitula 1-1.5 
cm diam. (radiis inclusis); involucri 2-seriati 
phyllaria lineari-lanceolata acuminata minute 
glandulari-pubescentia et sparsissime hirsuta, 
margine angusto scarioso leviter lacerato; 
radii albi; pappus minutus coroniformis vel 

annularis laceratus cum setis paucis caducis 
interioribus; achenia basi late callosa dense 
pubescentia pilis longis apice rectis. 

Jalisco: Canyons, moist rocks, Tuxpan, 
February 1904, Pur-pus 527 (G, type). 

The serrate leaves, crowded coombs, and 
cinereous aspect are characteristic. 

Achaetogeron garciae (Blake) 
Larsen, comb. nov. 

Bellis garciae Blake, Contr. U. S. Nat. Herb. 
22:593. 1924. 

Still known only from the type, P. Ibana 
Garcia 310 (U. S. Nat. Herb. 1032782) from 
the State of Durango, alt. 100 m. 

ENTOMOLOGY.' — Notes and descriptions of Near die Hydroptilidae (Trichoptera) } 
Herbert H. Ross, Illinois Natural History Survey, Urbana, 111. 

In the caddisflies of the family Hydroptil- 
idae there undoubtedly exists a number of 
genera that bear their closest relationship 
with forms known from other regions and 
with which they have not yet been associ- 
ated. Two such cases are outlined in this 
paper, and in each the geographic limits of 
the genera involved are extended into 
another continent. During the investigation 
of these cases some interesting morpho- 
logical considerations have arisen, and as a 
result a few suggestions regarding termi- 
nology are made in this paper. 

I am greatly indebted to Dr. G. T. Riegel, 
University of Illinois, and to Dr. H. K. 
Gloyd, Chicago Academy of Sciences, for 
making material available to me for study, 
and to Dr. D. G. Denning for the loan of 
type material of some of the species studied. 
Types of new species described in this paper 
are deposited in the collection of the Illinois 
Natural History Survey. 


In reviewing the terminology I have used 
for parts of the male genitalia of Hydroptili- 
dae, one change and one addition seem indi- 
cated in the interests of clarity and struc- 
tural identification. 

Subgenital plate. — In most genera of Hy- 
droptilidae there is a variously shaped, 

1 Received March 8, 1948. 

mesal structure usually occurring above or 
between the claspers. In Hydroptila, Fig. 6, 
sp, and Neotrichia this structure is platelike 
and bears at its apex a pair of setae. In 
Oxyethira and Stactobiella what appears to 
be this same structure is heavily sclerotized, 
frequently arched, and assumes a variety of 
shapes in different species; in these two 
genera I have heretofore called this struc- 
ture the tenth tergite (Ross, 1938, 1944). 
The tenth tergite, however, must be above 
the aedeagus, whereas in the structure in 
question, Fig. \A, sp, it is situated below the 
aedeagus. It may be a sort of guide for the 
aedeagus or it may assist in the coupling 
action during copulation. Because it lies 
below the aedeagus and because its exact 
homology is obscure, I am proposing that 
this plate be called the subgenital plate. 

Bracteole. — In quite a number of Hy- 
droptilidae, especially in Stactobiella, there 
occurs a structure associated with the area 
dorsad of the base of each clasper. In some 
cases, Fig. 1, br, this appears as a small 
structure at the base of each clasper, in 
others, Fig. 2, br, the structure is larger and 
more conspicuous than, the clasper and 
probably usurps its function. For this I 
propose the term bracteole. Here again the 
homology is difficult to determine, but I 
believe that the term will be a useful one for 
purposes of taxonomic description. 



Genus Stactobiella Martynov 

Stactobiella Martynov, Pract. Ent. 5: 58. 1924. 

(Genotype, monobasic, Stactobia ulmeri Sil- 

Tascobia Ross, Bull. Illinois Nat. Hist. Surv. 23: 

124. 1944. (Genotype, by original designation, 

Stactobia palmata Ross.) New synonymy. 

Comparing Nearctic species of this genus 
with illustrations of Palearctic species shows 
that not only are the two genera synonymous 
but also that two Nearctic forms each have 
their closest known relative in the Palearctic 
fauna. S. palmata (Ross) is very similar to 
Martynov's illustrations of biramosa Marty- 
nov, the former having the claspers short and 
ovate, constricted at base, and the stalked 
process tridentate, the latter having the clasp- 
ers slightly longer and parallel-sided, and the 
stalked process bidentate. In each the aedeagus 
is simple and tubular. The Nearctic species 
delira (Ross) is similar to the Palearctic ulmeri 
(Siltala) and risi (Felber), especially in regard 
to the curved, fmgerlike subgenital plate, but 
exhibits marked differences in the shape of each 
structure. The Nearctic brustia Ross is not 
allied closely to any of the other species and 
forms a species complex of its own. Thus, not 
only is Stactobiella as a genus Holarctic in dis- 
tribution, but two of its three component 
phylogenetic units are truly Holarctic also. 

To assist in identifying these species, a key 
is given to the males of these six species, which 
comprise the known world fauna of the genus. 
I have taken characters of the Palearctic forms 
from illustrations in the literature, cited in the 


1. Claspers apparently fused to form a ventral 

plate bearing three whiskerlike brushes of 
setae; aedeagus with apex curved and bent 
into shape of a crook (Ross, 1938, p. 115, fig. 
22). Nearctic (Wyoming) . . . .brustia (Ross) 
Claspers not fused, either elongate or biscuit- 
shaped 2 

2. A curved process, the bracteole, arising above 

each clasper, longer than the clasper and di- 
vided at apex into two or three fingerlike 
branches, Fig. 2; subgenital plate short and 

wide 3 

Bracteole represented by only a small process 
associated with a lateral clump of setae, 
clasper long, parallel-sided or tapering to 
apex; subgenital plate long and fingerlike, 
often curved or angled, Fig. 1 4 

3. Bracteole having apex divided into two 

"fingers" (Martynov, 1934, p. 159, fig. 105). 

Palearctic (Russia) biramosa Mar tynow 

Bracteole having apex divided into three 
"fingers," Fig. 2. Nearctic (central U. S. A.) 
palmata (Ross) 

4. Apical portion of aedeagus divided into one 

mesal and two lateral lobes, Fig. I A. Nearc- 
tic (northcentral U. 8. A.) delira (Ross) 

Apical portion of aedeagus tubular, much as in 
Fig. 2 5 

5. Apical margin of clasper truncate and slightly 

oblique (Felber, 1908, p. 721, fig. 2). Palearc- 
tic (Switzerland) risi (Felber) 

Apical margin of clasper rounded, the clasper 
tip narrow (Martynov, 1934, p. 159, fig. 

103). Palearctic (northern Europe) 

ulmeri (Siltala) 

Genus Oxyethira Eaton 

Argyrobothrus Barnard, Trans. Roy. Soc. South 
Africa 21: 392. 1934. (Genotype, monobasic, A. 
velocipes Barnard.) New synonymy. 

Barnard's illustrations of pupal case, vena- 
tion, and genitalia of both sexes leave no 
doubt that Argyrobothrus velocipes is a typical 
member of Oxyethira. The genus as such has 
not been recorded previously from central or 
southern Africa. 

Oxyethira arizona, new species 

The wide, curved lateral process of the aede- 
agus and the long, ventral internal projection 
of the base of' the ninth segment show a rela- 
tionship of this species with pallida (Banks) 
and maya Denning, but arizona differs from 
both of these in the straight apical portion of 
the aedeagus (this part is angled and twisted 
in pallida and maya) and the elongate lateral 
processes of the eighth tergite. 

Male. — Length from front of head to tip of 
wings, 2.5 mm. Color entirety pallid or straw 
color, with an annulation of a darker shade on 
some segments of the antennae and some darker 
areas on the front wings. General structure 
typical for genus. Seventh sternite with a sharp 
curved apical process. Genitalia as in Fig. 4. 
Eighth tergite divided into a pair of lateral 
lobes, the lower margin of each produced into 
a Jong, smooth, sharp process, the left one 
curved dorsad at tip, as shown, the right one 
curved slightly ventrad; the base of each lobe 
bears abundant long setae, but the projecting 
portion is without vestiture. Eighth sternite 
forming the lateral and ventral part of a ring, 
trapezoidal from lateral view, the apical mar- 
gin gently curved on the meson. Ninth tergite 
membranous. Subgenital plate moderately 

Fig. 1. — Stactobiella delira, male genitalia, ventral aspect; IA, same, lateral aspect. Fig. 2. — 
Stactobiella palmata, male genitalia, ventral aspect. Fig. 3. — Oxyethira lum-osa, male genitalia, lateral 
aspect; 3A, apical portion of male genitalia, ventral aspect. Fig. 4. — Oxyethira arizona, male genitalia, 
lateral aspect; 4A, claspers and associated structures, ventral aspect; 4J5 aedeagus; 4C, aedeagus as 
seen at right angles to long axis of 45. Fig. 5. — Neotrichia numii, male genitalia, lateral aspect; 5A, 
aedeagus, 5B, claspers and associated structures, ventral aspect; 5C, bracteole. Fig. 6. — Hydroptila 
denza, male genitalia, ventral aspect; 6A, aedeagus. Abbreviations: br, bracteole; cl, clasper; sp, 
subgenital plate. 



VOL. 38, NO. 6 

heavily sclerotized, lateral aspect evenly curved 
to form about a quarter of a circle, ventral 
aspect, Fig. 4 A, incised at apex to form a mesal 
arcuate area on each side of which is a short 
fingerlike process. Claspers fused on meson, 
very deep dorsoventrad, the ventral apical 
margin bearing a fingerlike projection on each 
side with a truncate edge between them. Style 
sinuate, membranous, and tipped with a long 
seta. Aedeagus, Figs. 4B and C, with base only 
a little longer than apex; neck only indistinctly 
set off: and apex divided at base into a taper- 
ing, slightly twisted central portion bearing 
the penis, and a wide, ribbonlike twisted pro- 
cess which makes a complete circle around the 
central portion, and which is concave on its 
inside surface. 

Female. — Size, color, and general structure 
similar to male. Genitalia simple; tenth tergite 
fairly wide at base, tapering to a round apex; 
ninth tergite with a narrow, spiculate, dorso- 
apical hump and with straight internal apo- 
demes. Spermatheca and its associated struc- 
tures similar in most respects to pallida, but 
differing in having the ventral bar of the sperma- 
theca arcuate but only moderately wide. 

Holotype, male. — Superior, Pinal County, 
Ariz., taken at light in Boyce Thompson 
Arboretum, May 17-24, 1946. 

Allotype, female. — Same data as for holotype. 

Paratypes. — Same data as for holotype, 2 d\- 

Oxyethira pallida (Banks) 

Oxyethira cibola Denning, Can. Ent. 79: 12. 1947. 
New synonymy. 

The most outstanding character of this 
species is the more slender of the two curved 
lateral processes of the aedeagus. This is ab- 
sent in maya Denning; the other apical struc- 
tures of the aedeagus are subject to twisting in 
both species and, in comparably twisted speci- 
mens, are remarkably similar in both. 

Collecting generously afforded by Dr. Swingle 
around artificial fish ponds near Auburn, Ala., 
brought in several catches of this species. It is 
probable that it is the first species of caddisfly 
to invade these ponds after they are filled. 

Oxyethira lumosa, new species 

This species is most closely related to grisea 
Betten and novasota Ross on the basis of simple 
eighth segment and the long, curled spiral 

process of the aedeagus. From both species and 
others in the genus it differs markedly in the 
small, compact, ovate ventral aspect of claspers 
and subgenital plate, Fig. 3A. 

Male. — Length from front of head to tip of 
wings, 2.5 mm. Color a salt and pepper mixture 
of cream and brown. General structure typical 
for genus. Seventh sternite with a sharp apico- 
mesal spur. Genitalia as in Fig. 3. Eighth 
segment almost cylindrical, the apical margin 
slightly roughened and its ventral aspect U- 
shaped. Ninth segment with ventral portion 
forming only a rounded internal lobe extending 
only slightly into the seventh segment. Ninth 
segment also having no prominent marginal 
structures, the dorsal portion membranous and 
the ventral portion membranous and emargi- 
nate to form a deep V around the base of the 
claspers. Subgenital plate with lateral aspect 
forming a stout, hook-shaped structure with a 
wide base and stout, curved apex with a sharp 
tip, and with ventral aspect ovate and slightly 
incised on meson. Claspers fused to form an 
ovate plate deeply incised on meson from apex 
to over half the distance to base, Fig. 3A. 
Aedeagus not very long, base only about as 
long as apex, and the neck indicated by a shal- 
low constriction and the origin of the spiral 
process. This structure is stout and long, and 
encircles the aedeagus one and a half times, 
the first circle making almost a ring and the 
other half circle made while extending pos- 
teriad to the apex of the aedeagus. Apex of 
aedeagus cylindrical, semimembranous, and 
with a sclerotized, sharp, triangular sclerite 
placed transversely across the structure near 
the tip. 

Holotype, male. — Daytona Beach, Fla., 
August 27, 1945, G. T. Riegel. 

Genus Hydroptila Dalman 

This is the most abundant and widespread 
genus of Hydroptilidae, with species known 
from every part of the globe. It is interesting 
that the Nearctic fauna contains many species 
whose closest relatives are in the Palearctic 
fauna, and other species which have apparently 
arisen from Neotropical nuclei of evolving forms. 
The species described below is of this latter 

Hydroptila denza, new species 
Among some material from Mexico a speci- 

June 15, 1948 



men was encountered that is of the general 
type of meralda Mosely, but differs from it in a 
variety of characters: The claspers are little 
longer than the subgenital plate, the apex of the 
tenth tergite is deeply incised, and the apical 
blades of the aedeagus are wide and swordlike, 
Fig. 6A. 

Male. — Length from front of head to tip of 
wings, 3 mm. Color moderately dark brown, 
the wings mottled with gray and brown. 
General structure typical for genus and scent 
cap ovate. Seventh sternite with a short, sharp 
apicomesal process. Genitalia as in Fig. 6. 
Ninth segment projecting freely from eighth 
segment, its internal portion rounded and only 
moderately produced, the apicolateral margin 
bearing a fingerlike lobe which appears sharp 
and spurlike from ventral view. Tenth tergite 
with lateral margins sclerotized, mesal portion 
membranous and deeply incised. Clasper of 
moderate length, ventral aspect with apex 
slightly widened, lateral aspect with apex 
much widened and trianguloid; apicolateral 
corner projecting as a small sharp point, apico- 
mesal corner slightly angulate and a small 
sclerotized point just within it. Subgenital 
plate triangular, over two-thirds as long as 
clasper, and bearin'g a pair of setae near apex. 
Aedeagus, Fig. 6 A, with base extremely long, 
extending internally through three full seg- 
ments of a completely extended specimen; neck 
bearing a rufflike collar of membranous folds 
and a slender spiral process which encircles 
the aedeagus one and a half times; base of apex 
swollen, the apex beyond this divided into two 
parts, a long, bladelike, flattened, sclerotized 
process, and a tapering, straight style which 
bears the penis tube and which is membranous 
and corrugated at its base. 

Holotype, male. — Hacienda Santa Engracia, 
Tamaulipas, Mexico, March 9, 1939. 

Genus Neotrichia Morton 

As is true of Ochrotrichia, Mayatrichia, and a 
few other genera, additional United States 
species of Neotrichia continue to follow patterns 
of general structure found in Central American 
forms. To date these genera are known only 
from the New World, and their distribution 
seems to indicate clearly that these genera 
originated in the Neotropics, and various spe- 
cies are spreading northward following the 
retreat of the glacial area. 

Neotrichia numii, new species 
The elongate and triangular claspers and the 
structure of the aedeagus indicate definite 
affinities between this species and digitata 
Mosely on the one hand and collata Morton 
on the other. From both this species differs in 
the extremely long, beaklike process of the 
subgenital plate, and the very dissimilar sclero- 
tized processes at the apex of the aedeagus. 

Male. — Length from front of head to tip of 
wings, 2.5 mm. Color entirely dark brown, the 
legs paler. General structure typical for genus. 
Genitalia as in Fig. 5. Ninth segment having 
an angulate internal portion, each side ending 
in a spurlike process; the dorsal portion is 
fused with the tenth tergite, the lateral portion 
is large and its apical margin is produced into 
a low, wide angle. Tenth tergite moderately 
narrow, rounded at apex and forming a simple, 
undivided mesal structure projecting above the 
other apical parts of the genitalia. Clasper 
elongate, lateral aspect tapering and curved at 
apex, ventral aspect having a broad, parallel- 
sided base narrowing suddenly to a short, 
pointed apex, heavily sclerotized and bearing 
only a few short setae. Bracteole, Fig. 5C, 
spatulate, pale, and bearing a series of long 
hairs on its ventral and apical margin, the 
base of the bracteole short and narrow, the 
apical portion broad and slightly curved dor- 
sad. Subgenital plate unusually complex; the 
mesal portion, Fig. 55, ends in a long beak 
which in lateral view is narrow and sharp, and 
in ventral view is expanded at apex into a 
platelike area bearing a hooked lateral process 
and a pair of mesal setae; this mesal part is 
joined ventrad with a large, convex lateral 
sclerite which narrows dorsad and appears to 
fuse with the inner margin of the tenth tergite. 
Aedeagus, Fig. 5A, elongate, the base tubular 
and narrowing to a long neck from the apex 
of which arises a stout spiral process encircling 
the aedeagus a little more than a complete 
turn; apex swollen at base, and divided at 
apex into a sharp spur bearing the penis and a 
lateral plate that is broad at base, slightly 
narrower at apex, and almost truncate at tip. 
Holotype , male. — Lake George, Colo., in 11- 
mile canyon of the South Platte River, August 
8, 1943, J. A. and H. H. Ross. 

Neotrichia vibrans Ross 
Neotrichia ranea Denning, Can. Ent. 79: 20. 1947. 
New synonymy. 



Further study of this species indicates that 
there is considerable lateral movement of 
several parts of the genital capsule, and that 
this may result in considerable difference of 
appearance between one specimen and another. 
The widely expanded condition is shown in 
my own drawing (Ross, 1938, p. 120, Fig. 29), 
and a more contracted condition is illustrated 
by Denning in the description of ranea. Dr. 
Denning has kindly loaned me his type ma- 
terial for study. 


Barnard, K. H. South African caddis-flies 
(Trichoptera) . Trans. Roy. Soc. South 
Africa 21: 20-394, illus. 1934. 

Denning, Donald G. Hydroptilidae (Tri- 
choptera) from southern United States. 
Can. Ent. 79: 12-20, illus. 1947. 

Felber, Jacques. Microptila risi nov. sp. 
Zool. Anz. 32: 720-722, illus. 1908. 

Martynov, Andreas B. Trichoptera. Prac- 
tical Entomology 5: 67 and 388 pp., illus. 
1924. Leningrad. 

. Trichoptera Annulipalpia. Tabl. 

Analyt. Faune U.R.S.S. 13: 1-343, illus. 

Ross, Herbert H. Descriptions of Nearctic 
caddis flies {Trichoptera) with special refer- 
ence to *the Illinois species. Bull. Illinois 
Nat. Hist. Surv. 21 (4): 101-183, illus. 

. The caddis flies, or Trichoptera, of 

Illinois. Illinois Nat. Hist. Survey 23 
(1): 1-326, illus. 1944. 

ZOOLOGY. — An analysis of specific homonyms in zoological nomenclature} 
Richard E. Blackwelder, U. S. National Museum. 

In the systematic study of animals, 
which is the science known as systematics 
or taxonomy, the scientific names of the 
animals are at once an essential tool and a 
source of much confusion and discussion 
because of their vast number and the com- 
plexity of our system of using them. A 
branch of systematics has grown up which 
concerns itself entirely with these names 
and the principles to be used in applying 
them; this is nomenclature. Its principal 
aims are to formulate and establish systems 
under which each species can be given 
a distinctive name and to provide machin- 
ery to insure as great stability or perma- 
nence as possible to each of these names. 

One of the difficulties which plague the 
taxonomist in his use of scientific names is 
the situation that arises when the same 
name is inadvertently given to two different 
species of animals. If a name is to be useful 
in exact science it must always refer to but 
one species, and always to the same species. 
Therefore, we cannot permit the use of one 
name for two or more species, and when 
duplicate names are found we must provide 
another name for one of the species. Identi- 
cal names used for two or more species are 
called homonyms. They may be further 
classified by calling the older of the two 
usages the senior homonym and the younger 
one the junior homonym. 

1 Received April 2, 1948. 

The discoveiy of homonyms very often 
results in a change of name for one of the 
species, and this type of change accounts 
for a large proportion of the annoying al- 
terations of names that have given taxon- 
omy a bad reputation among biologists. It 
is thus of special importance to taxono- 
mists to study the problem of homonymy 
and find a way to protect names from this 
major source of instability. 

The treatment of homonymy of specific 
names in zoological nomenclature is one of 
the oldest problems with which the writers 
of rules of nomenclature have had to deal, 
and it is one that has not been solved on 
any universal basis even after 75 years of 
code-building. It is therefore not unreason- 
able to reexamine the problem to present a 
possible new approach. 

In the various nomenclatural codes that 
have been proposed, the treatment of the 
problem of homonymy differs widely. The 
reason for this divergent treatment is not 
clear, except on the assumption that none 
of them have given real satisfaction. It is 
therefore believed useful to examine the 
procedures that have been proposed here- 
tofore and to analyze the requirements of a 
satisfactory system. 

(I) The first method we will examine for 
dealing with homonyms is that prescribed 
by the International Rules of Zoological 
Nomenclature in articles 35 and 36. It is 

June 15, 1948 blackweldeh: specific homonyms in zoological nomenclature 20' 

the most direct and the simplest approach 
but may not be the one giving the most 
desirable results. This treatment is based on 
two principles: (1) "A specific name is to be 
rejected as a homonym when it has pre- 
viously been used for some other species or 
subspecies of the same genus" (from article 
35), and (2) "rejected homonyms can never 
be used again" (from article 36). 2 This is a 

2 The term specific name is here used in its cur- 
rently official meaning, in which the combination 
of the generic and the trivial names is a specific 
name. Article 2 holds that the scientific designa- 
tion of a species is binomial, and it is therefore 
only the binomial combination that can be a 

laoo ALBUS fab 


simple and direct solution, and it has been 
reinforced by Opinion 83. Unfortunately 
the ramifications of nomenclature are not 
as simple as this rule, and names do not 
conform to a pattern of being irrevocably 
either "homonyms" or "not homonyms" 
at any given time. 

This can be illustrated by diagrams of the 
eight major types of homonyms (Fig. 1). 
You may recognize in these diagrams your 
old friends X-as albus and Y-us albus. In 
these diagrams two genera are represented 
in parallel vertical columns. The passage 
of time as one reads down each diagram is 
indicated by the dates at the left, and the 













I I 



ALBUS LATR ' 1800 







Fig. 1 






subsequent history of each specific name Is 
indicated by the solid line beneath it. The 
heavy horizontal line at the bottom repre- 
sents the time at which the homonymy is 
discovered. In A, albus was proposed twice 
in X-us, producing unequivocal homonymy 
from 1900 on. B is just like A except for the 
later removal of one of the species to Y-us. 
It would be just the same if it had been the 
other name that was removed. In C the 
older name has been removed before the 
proposal of the second name. In D the two 

were originally proposed in separate genera, 
but by the time the second was proposed 
in X-vs, the older name had been trans- 
ferred to X-us, producing unquestiona- 
ble homonymy. E is similar except that 
the transfer of the older name is later 
than the proposal by Smith. And F, G, H 
are the same as D and E except that the 
transfer of the older name to X-us was 
temporary — in F entirely prior to Smith, in 
G entirely subsequent to Smith, and in H 
partly before and partly after Smith. 























Fig. 2 

June 15, 1948 blackwelder: specific homonyms in zoological nomenclature 209 

Under the International Rules every one 
of these eight types of homonyms requires 
a change of name, because in every case 
there has been combined with X-us a name 
albus that is older than X-us albus Smith 
1900. This is the rigid interpretation of the 
present wording of the Rules, which has 
been followed by many taxonomists but by 
no means by all. 

Many taxonomists have felt that in some 
of these cases, such as diagram F in which 
the homonymy was of a tempoiary histori- 
cal nature, it is not really necessary to re- 
place the 1900 name. They have argued that 
a distinction of some sort should be made to 
prevent the change of such names. 

(II) The question of whether all homo- 
nyms should be treated alike has led to wide 
discussion of a possible distinction between 
two types called primary and secondary 
homonyms, so that different treatment might 
be accorded them after revision of this part 
of the Rules. There have been at least two 
definitions of primary and secondary homo- 
nyms, but the one most commonly known 
is this. A homonym is primary if the 
names were originally proposed in the 
same genus; it is secondary if the two names 
occur in the same genus only through trans- 
fer of one from another genus. It is argued 
that all primary homonyms must be re- 
placed, but that secondary homonyms 
should require replacing only if the names 
are still in the same genus. The examples in 
the top row of Fig. 1 (A, B, C) are primary, 
since the names were originally in the same 
genus; all the rest are secondary, under this 
definition, because the names were origi- 
nally in different genera. A, B, C require 
replacing of the younger name, since these 
are primary homonyms, but D and E also 
require replacing, because both names are 
currently in use in X-us. Only F, G, and H 
can be saved by this procedure. Stability 
of the names is indicated by a tabulation, 

Method Change Keep 

I. International Rules 8 

II. Primary-Secondary (original 

genus) 5 3 

(III) A third school employs primary and 
secondary in a very different sense. Here, a 
homonym is primary if it was a homonym 

at the time of its proposal, whether this 
fact was recognized or not; it is secondary 
if it was valid when proposed and became 
a homonym later by other action. 

Again, the primary homonyms are treated 
as in the International Rules — they must 
be replaced. And secondary homonyms are 
considered to be such only as long as the 
two names remain in the same genus. In 
Fig. 2 we find the top two rows representing 
the primary homonyms under this defini- 
tion. A, B, C, D, F, and H are primary, 
since the later name was already preoccu- 
pied by an older combination with X-us. 
But E, which is secondary, must also be 
replaced, since the names are both in cur- 
rent use in X-us. This would appear in the 
stability tabulation, thus : 

Method Change Keep 

I. International Rules 8 

II. Pri-Sec (original genus) 5 3 

III. Pri-Sec (validity when pro- 
posed) 7 1 

(Ilia) A variant of the third procedure 
might be required by those persons who 
believe that temporary transfer of a name 
into X-us does not preoccupy the name for 
later use in X-us. In this case, A, B, C, D, 
and E would be primary. This would give 
results in the stability table of Change 5, 
Keep 3. 

The first of these procedures for separat- 
ing homonyms into primary and secondary 
(our second system) is based on two new 
principles (see Fig. 1) : (1) The fact that the 
identical names were originally proposed 
in the same genus is held to be of first im- 
portance, making the top row primary. In 
this manner, in Fig. 3, diagram A is set 
apart from diagram D, although they are 
identical except for the original assignment 
of the names, yet we must take the same 
action in the two cases because there is 
actual homonymy in each. They must be 
treated alike even though one is primary 
and one secondary. In diagrams B and H, 
which also are identical except for the orig- 
inal genus of the names, we should be re- 
quired to change a name in B and not in H 
because B is primary and H is secondary. 
This time we do not treat them alike, al- 
though they differ in exactly the same way. 
There appears to be no justification for 




,*» ALBUS 



*00 ALBUS 9-rrvi 








tseo ALBUS smtw 



these distinctions, and there also appears to 
be no basis for the principle of segregating 
homonyms on the basis of their original 
assignment. (2) A secondary homonym must 
be replaced only if the older name is still in 
the same genus at the time the homonym is 
discovered. There can be no question that 
the younger name in diagram D must be 
replaced, because there is active homonymy 
at the present time, but if in diagram H it is 
not necessary to rename the younger name, 
why must we do so in diagram B? Merely 
because of the accident of the original gen- 
eric assignment of the older name? There 
appears to be no justification for the use of 
this principle either. 

The second procedure for separating 
homonyms into primary and secondary uses 
another new principle (see Fig. 4). It places 
first importance upon the fact that Smith in 
1850 (in E) was correct in believing that 
the name albus had never been used before 
in the genus X-ns, whereas in D he failed 
in 1900 to recognize the older name. D is 
therefore called primary and E secondary. 
But in these cases if the older name had been 

later removed from X-us (H and G, in 
which H is called primary and G secondary) 
we would be required under this third pro- 
cedure to replace the newer name in H 
(because it was a primary homonym) and 
would not be required to replace it in G 
(where it is secondary). Again we find that 
D (primary) and E (secondary) must be 
treated alike, whereas H (primary) and G 
(secondary) must be treated differently. 
There appears to be no reason for such a 

In summarizing these last two proce- 
dures, it is apparent that there has appeared 
the new principle that a name may not be 
replaced unless the older name is still in the 
same genus when the homonymy is dis- 
covered, as in D and E of Fig. 4. But this 
principle is applied only to the class of 
secondary homonyms as variously defined. 
It would require a change in D and in E, 
and it would require a change in H but not 
G. There does not seem to be any reason 
why all types of homonyms should not be 
treated in this way. In other words, if there 
is no real justification for the distinction 





ALBUS L«m *oo ALBUS un. 

*^o ALBUS s-th 


Fig. 4 

June 15, 1948 blackwelder: specific homonyms in zoological nomenclature 211 

leoo ALBUS fab 



i85o ALBUS smith 









isoo ALBUS SMrm 








ALBUS um. 


Fig. 5 

between primary and secondary, why not 
judge both types on the same ground? 

(IV) Now if we apply this new plan uni- 
formly to all types of homonyms, we need 
new terms for the ones that are to be re- 
placed and those that are not, to avoid con- 
fusion with primary and secondary. We are 
now interested, in the first place, in the situ- 
ation as we find it at the time of discovery 
of the homonymy. (This time is represented 
by the heavy line across the bottom of each 
diagram.) The question to be asked is, Is 
there concurrent use of two names of iden- 
tical spelling? 

In Fig. 5, in A, D, and E two identical 
names are in use in the genus X-us at the 
present time — they occur together and are 
therefore coincident homonyms. In B, C, F, 
G, and H the names are not now in the 
same genus and are therefore homonyms 
only in an historical sense — only because 
of their antecedents. We may therefore 
term them antecedent homonyms. 3 

3 Other terms may be thought more readily 
understood in these uses, such as concurrent and 
historical or present and past, but the need for 
any such terms would be eliminated in a new rule 
by calling only the first group "homonyms," 
eliminating the need for any term for the second 



By replacing only the coincident homo- 
nyms, we obtain a rating on the stability 
tabulation thus: 





Intcinational Rules 



Pri-Sec (original genus) 
Pri-Sec (validity when pro- 











I should point out here that in this table 
a very false impression can be made. These 
figures are the number of types of homo- 
nyms that require change. One of these 
types might be much more common than 
another, completely overshadowing it in 
importance. But it appears that in number 
IV, changes are required only in cases which 
would have had to be changed under all of 
the other systems as well (A, D, E in Fig. 
5), and the remaining cases are an improve- 
ment in stability over the other procedures. 

A summary of this distinction between 
coincident and antecedent homonyms can 
be made by defining them and the treat- 
ment of them. Any name that is discovered 
to be of the same spelling as an older name 
currently assigned to the same genus is a 
and is to be replaced. Any name that is dis- 
covered to be of the same spelling as an older 
name that was at one time in the same genus 
but is not now so assigned is an ANTE- 
not to be replaced. 

Obviously, an antecedent homonym can 
become coincident, as would be the case in 
diagram Cif after 1900 the older name was 
brought back into X-us. It is the actual 
state of affairs at the time of discovery of 
the homonymy that determines the type of 
homonymy and the action to be taken. 

In some discussions of primary and second- 
ary homonymy there has been proposed 
another new principle, which is a radical 
departure from the International Rules in 
that it requires the revival of a name pre- 
viously suppressed as a homonym. 

In Fig. 6, which represents a case of 
secondary homonymy in procedure II, 
X-us albus Smith, 1900, required replacing, 

even though it was a secondary homonym. 
It was renamed in 1910 by Jones as X-us 
novus. This is an action that would have 
been necessary under any of the four pro- 
cedures we have examined. Then in 1920 
albus of Latreille, the older name, was re- 
moved by Brown from X-us; the new pro- 
cedure would then require that X-us albus 
Smith 1900 be reinstated with novus Jones 
as a synonym. 

This appears to be completely contrary 
to the primary aim of the Rules, which is 
stability in nomenclature, since it produces 
an extra name change that is not absolutely 
necessary. If it stopped here, it would be 
fairly reasonable, but there is nothing to 
prevent some other worker from claiming 
(in 1925 for example) that albus of Latreille 
was correctly transferred to X-us, and 
therefore novus must be used for albus 
Smith once again. And in 1926 Brown reit- 
erates his belief that albus Latreille does not 
belong in X-us, and therefore novus is only 
a synonym of X-us albus Smith. 

It is more than possible that such a con- 
troversy should continue for years, since 
there are many cases in which two authors 



»oo ALBUS si*™ 

wo NOVUS *»*s 

Fig. 6 

June 15, 1948 

proceedings: the academy 


have consistently used different generic 
assignments for a given species. 

Again, when two workers disagree on the 
correct generic assignment of a species, as 
has occurred many times over long periods 
of years, the first worker, calling the genus 
X-us, finds that X-us alb us is a secondary 
homonym, as in D, requiring a new name; 
whereas simultaneously and equally cor- 
rectly the second worker, who believes the 
genus should be called Z-us, finds that the 
name Z-us albus is not preoccupied at 1900 
and is therefore the correct name for the 
species. On the basis of their beliefs concern- 
ing the generic name, both are correct in 
their use of the specific names, so that we 
have two apparently correct names in use 
for one species. 

The possibility of this result from a dif- 
ference of opinion on generic assignment of 
another species, or of the correct name for 
a genus, can not occur under our present 
Rules. In procedure II as generally stated 
it can occur in five of the examples. In 
procedure III it can occur in two of the 
examples. If the conclusions of this discus- 
sion are made part of procedure IV, such a 
situation could not arise in anv case. 

The only point that I have seen raised 
against this principle of accepting a valid 
renaming in spite of later actions is that 
malicious transfer could provide the means 
for some one to rename any species desired, 
on a small or large scale. I have not found 
any one who knows of such a case of mali- 
cious misappropriation of a rule, and I con- 
sider it at best as a possibility of importance 
insufficient to warrant action designed to 
prevent it. If a case should occur, the 
International Commission has ample au- 
thority to take action against the offender, 
without adding to the complexity of our 
nomenclatural system. 

There appears to be no other reasonable 
course than to accept the valid renaming 
of a homonym as final. There are other 
cases in which we find that strict priority 
gives less stability to names than a care- 
fully controlled departure from priority. 
And this appears to be an opportunity to 
prevent the possible (or rather certain) con- 
fusion of having two valid trivial names in 
current use for one species, merely through 
a difference of opinion concerning the status 
of some other species which happens to 
have an identical name. 



The 50th annual meeting, concurrently with 
the 353d meeting of the Academy, held in the 
Auditorium of the Cosmos Club, January 15, 
1948, was called to order at 8:15 p.m. by the 
President, Waldo L. Schmitt, with 55 persons 
in attendance. 

The minutes of the 49th annual meeting were 
approved as published in the Journal 37: 436- 
444. 1947. 

The reports of several officers and of the 
Committees of Auditors and Tellers were read 
and accepted. These reports are recorded at the 
end of the minutes. 

After the acceptance of the report of the 
Committee of Tellers, the President declared 
the following duly elected to the given offices: 

Frederick D. Rossini, President, 
C. Lewis Gazin, Secretary, 
Howard S. Rapplete, Treasurer, 
Francis M. Defandorf and William N. 
Fenton, Board of Managers to January 1951. 

The Secretary presented for the Affiliated 
Societies their nominations for Vice-Presidents 
of the Academy as follows: 

Philosophical Society of Washington — Walter 

Anthropological Society of Washington — Wil- 
liam N. Fenton 

Biological Society of Washington — John W. 

Chemical Society of Washington — Charles 

E. White 

Entomological Society of Washington — Carl 

F. W. Muesebeck 

National Geographic Society — Alexander 

Geological Society of Washington — William 


Medical Society of the District of Columbia — 

Frederick O. Coe 
Columbia Historical Society — Gilbert Gros- 


Botanical Society of Washington — Ronald 

Washington Section of the Society of American 

Foresters — William A. Dayton 



Washington Society of Engineers — Clifford 

A. Betts 
Washington Section of the American Institute 

of Electrical Engineers — Francis B. Silsbee 
Washington Section of the American Society of 

Mechanical Engineers — Martin A. Mason 
Helminthological Society of Washington — 

Aurel O. Foster 
Washington Branch of the Society of American 

Bacteriologists — Lore A. Rogers 
Washington Post of the Society of American 

Military Engineers — Clement L. Garner 
Washington Section of the Institute of Radio 

Engineers — Herbert Grove Dorsey 
Washington Section of the American Society of 

Civil Engineers — Owen B. French 

The Secretary was instructed by the mem- 
bers present to cast a unanimous ballot for 
these nominees. 

The President announced the recipients of 
the Academy's Awards for Scientific Achieve- 
ment for 1947 as follows: 

In the Biological Sciences, no award made. 

In the Engineering Sciences, Harry W. 
Wells, of the Department of Terrestrial Mag- 
netism of the Carnegie Institution of Washing- 
ton, in recognition of his distinguished upper- 
air research and organization of a world-wide 
network of ionospheric stations. 

In the Physical Sciences, Robert D. Hun- 
toon, of the National Bureau of Standards, in 
recognition of his distinguished service in the 
advancement of electronics and its applications 
to other sciences and to modern ordnance. 

The reports of three special committees were 
presented as a part of the program as follows: 
The Journal, by the Chairman of the Commit- 
tee, Raymond J. Seeger; National Science 
Legislation, by the Chairman of the Commit- 
tee, A. T. McPherson; A Junior Academy of 
Sciences, by the Vice- Chairman of the Com- 
mittee, Egbert H. Walker. 

The retiring President, Waldo L. Schmitt, 
presented his address, The Academy in Retro- 
spect and Prospect, then appointed Past Presi- 
dents Eugene C. Crittenden and Oscar E. 
Meinzer to escort the incoming President, 
Frederick D. Rossini, to the chair. Dr. 
Rossini adjourned the meeting at 10:20 p.m. 


During the Academy year, January 16, 1947, 
to January 15, 1948, one person was elected to 
honorary membership and 45 persons were 
elected to regular membership, including 42 

resident and 3 nonresident. Of these, 26 resi- 
dent and 2 nonresident qualified for member- 
ship. Ten resident members, elected to mem- 
bership January 12, 1948, were notified on 
January 13, but insufficient time has elapsed for 
their reply. Sixteen resident members and 4 
nonresident members elected to membership in 
the preceding Academy year qualified during 
the present Academy year just ending. The 
new members were distributed among the vari- 
ous sciences as follows: 9 in botany, 8 in 
physics, 5 each in zoology and entomology, 4 
each in chemistry and engineering, 2 each in pa- 
leontology, geology, anthropology, and plant 
pathology, and 1 each in biology, bacteriology, 
genetics, physiology, and ceramics. 

Twelve resident and 3 nonresident members, 
having retired from the gainful practice of their 
professions, were placed on the retired list of 
members to enjoy all the privileges of active 
membership without further payment of dues. 
The deaths of the following members were 
reported to the Secretary: 

Charles S. Butler, Bristol, Tenn., on October 

7, 1944. 
Georges Perrier, Paris, France, on February 

16, 1946. 
Edward A. Goldman, Washington, D. C, on 

September 2, 1946. 
Sir James Harwood Jeans, Dorking, Surrey, 

England, on September 16, 1946. 
Willis L. Jepson, Berkeley, Calif., on Novem- 
ber 7, 1946. 
Henry G. Avers, Washington, D. C, on Janu- 
ary 19, 1947. 
Howard S. Roberts, 2d, Washington, D. C, 

on January 30, 1947. 
Charles A. Browne, Washington, D. C, on 

February 3, 1947. 
W. P. Hay, Bradenton, Fla., on May 26, 1947. 
Hardee Chambliss, Washington, D. C, on 

June 1, 1947. 
Rudolf W. Glaser, Princeton, N. J., on Sep- 
tember 4, 1947. 
Robert H. Lombard, Worcester, Mass., on 

October 11, 1947. 
D wight W. Windenburg, Washington, D. C, 
on November 14, 1947. 

Sir James Harwood Jeans and General 
Georges Perrier were honorary members. 

On January 15, 1947, the status of member- 
ship was as follows: 

Regu- Re- Honor- n „ . , 

. . . Patron Total 


2 511 

15 202 




. 457 


Nonresident . . . 

. 161 







June 15, 1948 

proceedings: the academy 


The net changes in membership during the 
past year are as follows: 

Resident . . . 
Nonresident . 

+ 8 
+ 9 

Total +17 

- 2 







+ 6 


During the Academy year 1947 the Board of 
Managers held 9 meetings, with an average at- 
tendance of 18 persons. The following impor- 
tant matters were considered by the Board: 

1. Observance of the 50th Anniversary of the 
Academy. — The President appointed a commit- 
tee to look into the various ways in which the 
50th Anniversary of the Academy might be ob- 
served and make recommendations to assist 
the incoming 1948 Board in its selection and 
carrying out of the observance. It was antici- 
pated that this committee might render timely 
services inasmuch as the anniversary date, 
February 18, 1948, falls such a short time after 
the induction of the new Board of Managers. 

2. Preparation of an Illustrated Anniversary 
Edition of the "Red Book" or Directory. — The 
Board of Managers authorized the publication 
of an Anniversary Edition of the Red Book, 
which is to include photographs of members of 
the Academy together with as complete a list 
as possible of the membership of the various 
scientific societies affiliated with the Academy, 
and to include the constitutions, bylaws and 
current lists of officers of the Academy and 
affiliated societies. The Board authorized the 
expenditure of $3,000 for the publication of the 
directory, with a leeway of $600 to cover un- 
expected charges, such as proofreading, etc. 

3. Consideration of a Monograph. — A mono- 
graph on The parasitic birds of Africa, by 
Herbert Friedmann, was presented to the 
Board of Managers and recommended for pub- 
lication by the chairman of the Monograph 
Committee. Estimates for the cost of publica- 
tion were secured and the Board referred the 
manuscript to the Board of Editors for review 
and recommendation. 

During the Academy year, seven meetings of 
the Academy were held as follows: 

On February 20, 1947, Hugh L. Dryden, as 
retiring President, offered an address entitled 
Exploring the fundamentals of aerodynamics. 
Owing to unusually inclement weather the 
paper was read by title only. 

On March 20, 1947, the 1946 Academy 
Awards for Scientific Achievement were pre- 
sented to Waldo R. Wedel, U. S. National 
Museum, for work in the biological sciences; to 
Martin A. Mason, Beach Erosion Board, for 
work in the engineering sciences; and George 
W. Irving, Jr., Bureau of Agricultural and 
Industrial Chemistry, for work in the physical 

On April 17, 1947, His Excellency, Mr. 
Herman Eriksson, Minister from Sweden, 
contributed a few remarks on Alfred Nobel and 
the Nobel prizes, introducing P. W. Bridgman, 
of Harvard University, who addressed the 
Academy on High pressures and their effects. 

On May 15, 1947, James B. Sumner, Cornell 
University, addressed the Academy on The 
chemical nature of enzymes. 

On October 16, 1947, Edward A. Doisy, St. 
Louis University School of Medicine, addressed 
the Academy on Vitamin K. 

On November 20, 1947, I. I. Rabi, Columbia 
University, addressed the Academy on The 
hyperfine structure of the hydrogens and other 

On January 15, 1948, the meeting of the 
Academy was devoted to the business of the 
50th Annual Meeting and included the address 
of the retiring President, Waldo L. Schmitt, 
entitled The Academy in retrospect and prospect. 

The meetings were all held in the Assembly 
Hall of the Cosmos Club. 

C. L. Gazin, Secretary 


Your Treasurer submits the following re- 
port concerning the finances of the Washington 
Academy of Sciences for the year ended De- 
cember 31, 1947: 


Dues, 1945 $ 30 .00 

1946 75.00 

1947 3,167.00 

1948 70 .00 

$ 3,342.00 



Subscriptions, 1944 $ 6 .00 

1945 6.00 

1946 6.00 

1947 735.09 

1948 557 .84 

1949 3.38 1 ,314.31 

Reprints, 1946 $ 515.41 

1947 672.73 1,188.14 

Sales, 1947 178.07 

Interest & Dividends 1,319.11 

Directory 0.25 

Cash received in connection with the transfer from the 6 percent to the 3.6 percent pref. 

stock of the Potomac Electric Power Co 58.00 

Payment of balance of principal on the Chicago Railways Bond 750 .00 

Received from Amer. Assoc, for Adv. of Sci 479.00 

Photostats and postage 2.32 

Overpayment of dues 1 .00 

Total receipts, 1947 ■■ $ 8,632.20 

Cash balance, Jan. 1, 1947 4,784.25 

Total to be accounted for $13,416.45 


1946 1947 Total 

Secretary's Office $ 82 .80 $ 224 .95 $ 307 .75 

Treasurer's Office 21.30 236.78 258.08 

Subscr. Mgr. & Custodian 48.36 48.36 

Meetings Committee 40.25 327.10 367.35 


Printing & Mailing 333.83 3,451.27 3,785.10 

Illustrations 50.90 516.44 567.34 

Reprints 108.59 399.88 508.47 


Ed. Asst 20 .00 270 .00 290 .00 

Misc., postage, etc 5.34 33.08 38.42 

2— $1,000 Series G bonds 2,000.00 2,000.00 

Grants in aid 479.00 479.00 

1948 Red Book 292.00 292.00 

Set of Journals 100.00 100.00 

Share in projector for Cosmos Club 27.85 27.85 

Photostats 2.23 2.23 

Refund, overpayment of dues 1 .00 1 .00 

Totals $663.01 $8,409.94 $9,072.95 

Cash book balance, Dec. 31. 1947 4,343.50 

Total accounted for $13,416.45 


Balance as per cash book, Dec. 31, 1947 $ 4 ,343 .50 

Balance as per Amer. Sec. & Trust Co. 

Statement of Dec. 12, 1947 $4,972.67 

Receipts deposited since Dec. 12, 1947 397.69 

Total $5,370.36 

Checks outstanding, as of Dec. 31, 1947 

No. 1018 $ 5.41 

1140 5.00 

1201 25.73 

1202 518.36 

1203 . . 292 .00 

1204 168.55 

1205 10.81 

1206 1.00 1,026.86 $4,343.50 

June 15, 1948 proceedings: the academy 217 


Washington Sanitary Improvement Co. 

Certificate No. 434—100 shares at $10.00 $1 ,000 .00 

Certificate No. 435—100 shares at $10.00 1 ,000 .00 

Certificate No. 527— 8 shares at $10.00 80 .00 

Certificate No. 539— 1 share at $10.00 10 .00 

Certificate No. 582 — 200 shares at $10.00 2,000.00 $ 4,090.00 

Potomac Electric Power Co. 

Certificate No. TAO 1977— 40 shares 3.6 percent pref. at $50.00 2,000 .00 

City of New York — 3 percent (Transit Unification) Due — June 1, 1980 

Certificate No. D 20186 $ 500 .00 

Certificate No. C 71038 100 .00 

Certificate No. C 71039 100 .00 

Certificate No. C 71040 100.00 800.00 

First Federal Savings & Loan Ass'n. 

Investment account book 5 ,000 .00 

Northwestern Fed. Savings & Loan Ass'n. 

Certificate No. 1380 $4 ,500 .00 

Certificate No. 1441 500.00 5,000.00 

United States Government 

Series G Bonds— No. M 332990 G $1 ,000.00 

M 332991 G 1 ,000 .00 

M 332992 G 1 ,000.00 

M 332993 G 1 ,000 .00 

M 1808741 G 1,000.00 

M 2226088 G 1 ,000.00 

M 2982748 G 1 ,000.00 

M 4126041 G 1 ,000.00 

M 5141346 G 1 ,000.00 

M 5141347 G 1,000.00 10,000.00 

American Security and Trust Co. 

Savings Account 46 . 87 

Total $26,936.87 

Cash balance 12/31/47 4,343.50 

Total $31,280.37 

Total as of Dec. 31, 1946 $30,718.62 

Total as of Dec. 31, 1947 31,280.37 

$ 561.75 

At the close of business on December 31, 1947, there were 50 members of the Academy who were 
in arrears, as follows: 

No. of members In arrears for 

25 1 year 

11 2 years 

5 3 years 
2 4 years 

6 5 years 
1 6 years 

Total 50 

Probably at least half of these will regain the status of "member in good standing" eventually. 
However, the above tabulation shows plainly that a "house cleaning" is in order, and it is hoped that 
this work can be attended to early in 1948. 

Howard S. Rappleye, Treasurer. 




The accounts of the Treasurer of the Wash- 
ington Academy of Sciences for the year 1947 
were examined by your auditing committee on 
January 12, 1948. 

The Treasurer's report was checked and 
found to be in agreement with the records. All 
disbursements had been authorized and were 
found to be supported by vouchers and can- 
celled checks. The securities of the Society were 
inspected and found to be in agreement with 
the list given in the report and to have all 
coupons attached that are not yet due. 

Your Committee highly commends the 
Treasurer of the Academy upon the orderly and 
businesslike manner in which he has main- 
tained the fiscal records. The Academy con- 
tinues to be indebted to Mr. Rappleye for his 
conscientious and efficient performance of the 
arduous duties of his office. 

Frederick D. Rossini, Chairman 
Austin H. Clark 
Harold E. McComb 


Volume 37 of the Journal, for the year 1947, 
contained 448 pages in 12 issues, a gain of 16 
pages over Volume 36 of the previous year, in 
this respect more nearly approximating the 
ideal of 500 pages recommended by previous 
editors. There were 59 papers distributed 
among the various sciences as shown in the 
table below. 

These figures show an even poorer balance 
between the physical sciences and the biological 
sciences than the previous year. In 1946 about 
one-fourth of the pages of the Journal were 
devoted to physical sciences including geology, 
whereas in 1947 the papers in the physical 
sciences used only about one-tenth of the pages 
of the Journal. The disproportion reflects the 

production of the biologists and the activity of 
the freshman editor, Dr. Stone, who has been 
successful in procuring good papers. In 1947, 
moreover, the members of the Academy sub- 
mitted over 50 percent of the papers published, 
an improvement over 1946 when less than half 
of the papers were submitted by our own mem- 
bers. A considerably larger number of papers 
was received than the editors felt merited pub- 
lication in the Journal. A slight decline in the 
number of papers published this year is offset 
by the space given to several papers, of which 
at least one has been widely noticed in this 
country and abroad. Anthropology continues 
to contribute a fair share of papers, and papers 
from its constituent disciplines may be expected 
to be a substantial part of the Journal. Owing 
to the vigilance of Dr. Hoffman, who assumed 
responsibility for necrology, the number of 
obituaries procured was double that of 1946. 

The Journal and the new Board of Editors 
face the new year with confidence. Instead of 
the customary dearth of papers, the supply on 
hand will carry the Journal through the first 
quarter of the year. Among the papers on hand 
and at the printer's are several in the physical 
sciences which the editors gathered during 1947 
or which came to them through the activities of 
the Meetings Committee. 

The Journal faced a financial crisis in 
September, when the George Banta Publishing 
Co., through its representative, informed the 
Senior Editor, Dr. W. N. Fenton, that the 
costs of printing had increased to the extent 
that they had found it necessary to submit a 
new contract which, in effect, would increase 
the cost of printing the Journal by about 31 
percent. Dr. Fenton reported the matter to the 
Board of Managers, who referred the question 
to the Executive Committee, who, meeting 
October 13, 1947, recommended action, which 
the Board of Managers voted, October 20, to 



of papers, 



of pages, 



by pages, 



of pages, 



of pages, 





















Other (History of Science) 

Obituaries, Proceedings, etc 








June 15, 1948 

proceedings: the academy 


increase the budget for the Journal by $225 to 
meet the increase in cost of publication for the 
remainder of the year, and authorized Dr. 
Fenton to sign the new contract with the 
George Banta Publishing Co. for printing the 

The original 1947 budget provided $4,000 for 
printing and mailing, to which was added $225 
(mentioned above), $300 for editorial assist- 
ance, and $50 for miscellaneous office supplies 
and postage, bringing the total to $4,575. To 
this is to be added the charges to authors, which 
amounted to $936.38, making a grand total of 

The disbursements for the Journal have 

Printing and mailing $3,831.14 

Illustrations 532.95 

Reprints 473.64 

Editorial assistance 295 .00 

Office, miscellaneous 42 .31 

Total . . . $5,175.04 

To recapitulate, the total cost of the printing 
bills was $4,837.73. From this may be sub- 
tracted $936.38, the amount charged to authors 
for reprints, excess illustrations, and excess 
typesetting charges. Thus the net cost to the 
Academy of printing the Journal was 

It should be noted that on several items the 
Editors are well within their appropriation and 
that, despite the rise in printing costs at the 
end of the year, they were able to get by with- 
out the supplemental appropriation. 

The Board of Editors wishes to acknowledge 
the cooperation of the Board of Managers and 
the officers of the Academy. The Associate 
Editors met with the Board of Editors early in 
the year and worked out a procedure for han- 
dling manuscripts in special fields, for reading 
galley proofs, and for consultation, in all of 
which the Associate Editors have supported the 
Board of Editors ably and willingly. Dr. R. E. 
Blackweldbx supplied for Dr. Stone during 
the latter's field work in Alaska, and special 
thanks are due to Mr. Paul H. Oehser, our 
editorial assistant, for taking over administra- 
tive duties during the summer months while 
Dr. Fenton was away teaching. Mr. Oehser's 
very considerable specialized knowledge and his 
genuine interest in the continual improvement 
of the Journal far exceed the implication of his 

The Editors also wish to cite Mr. Irwin H. 
Wensink, Washington representative of the 
George Banta Publishing Co., for his help in 
maintaining the continuity and high standard 
of the Journal and for the courteous way that 
he conducted negotiations for a new contract 
for his company. We do not think it entirely 
flattery when he tells us that the Academy's 
editorial set-up has come to be a model which 
his company recommends to other learned 

William N. Fenton 
James I. Hoffman 
Alan Stone 

report of the custodian and subscription 
manager of publications 


Nonmember subscriptions in the United States 131 

Nonmember subscriptions in foreign countries 75 

Total 206 

Inventory of stock as of December 31, 1947: 

Reserve Sets of the Journal: 

Bound Volumes 1-29 and unbound Volumes 

30-37 lset 

Unbound Volumes 1-37 3 sets 

Unbound Volumes 11-37 6 sets 

Unbound Volumes 16-37 11 sets 

Miscellaneous numbers of the Journal 18,911 

Total numbers in the Reserve Sets 8,070 

Grand total of back numbers 26,981 

The only important change that has taken 
place during the year 1947 was the resignation 
of Frank M. Setzler as Custodian and Sub- 
scription Manager of Publications. Mr. Setzler 
was forced to hand in his resignation toward the 
end of this year because of his leaving the 
country for an extended field trip. To fill this 
vacancy, I was appointed at the 411th meeting 
of the Board of Managers held on October 20, 
1747. Mr. Setzler has admirably filled this 
position for five years, and during this tenure 
of office he has organized the operations and 
records in a most efficient manner, so that when 
I took over this position I found everything in 
perfect order and all information readily acces- 
sible. I feel that Mr. Setzler deserves to be 
highly commended for the excellent way in 
which he has run this office. 

Sales and expenditures: 

During the year no complete sets were sold, 
but the sale of individual volumes and numbers 



has continued, 24 volumes and 115 numbers 
having been sold. One complete set (Vol. 1-36) 
was purchased from one of the members of the 

Income from Sales 
Miscellaneous numbers $178.07 


Set of Journal Vol. 1-36 $100.00* 

Postage and office expenses 48 .36 

1947 Budgetary allotment $ 50.00t 

Enexpenditures for 1947 48.36 

Unexpended balance $ 1 . 64 

* Deducted from special fund earmarked for purchase of 
volumes and numbers needed to form complete reserve sets. 

t This does not include the additional $50 that was added 
to the budget allotment by vote of the Executive Committee 
for the purpose of carrying out a campaign to increase the 
present subscription list. This money was not expended in 

Harald A. Rehder, Custodian and 
Subscription Manager of Publications 


A total of 325 envelopes were delivered to 
the Secretary. The count of valid ballots 
showed the following officers were elected: 

President, Frederick D. Rossini 
Secretary, C. Lewis Gazin 
Treasurer, Howard S. Rappleye 
Board of Managers to January 1951, Francis 
M. Defandorf and William N. Fenton 

Anna E. Jenkins, Chairman 
Miriam L. Bomhard 
Regina Flannery 

report of special committee 

on legislation 


The Committee on Legislation, composed of 
J. E. Graf, W. W. Rubey, and A. T. Mc- 
Pherson, reported on recent and pending 
legislation in three phases of Government ac- 
tivity, each of especial interest and importance 
to science. Unusual possibilities for the ad- 
vancement of science were seen in the inter- 
national exchange of scholars and information 
provided by the Fulbright Foreign Scholarship 
Act (Public Law 584, 79th Congress) and by 
the proposed Information and Educational 
Exchange Act of 1948 (HR 3342, January 7, 
1948). Funds for the program under the former 
act are in foreign currencies derived from the 
sale of surplus war property abroad. The latter 
bill which was expected to be passed at an early 

date contains provision for the dollars needed 
to send American scholars abroad and will ex- 
tend the international exchange of scholars to 
countries in which there was no surplus prop- 

The second phase of legislative action re- 
ported on was that concerned with the future 
of the synthetic rubber industry. The Com- 
mittee felt that the most important considera- 
tion for future national security regarding 
rubber lay in establishing an adequate research 
and development program, entirely apart from 
any production program. The cost of such a 
program would be less even than the cost of 
maintaining the unused plants in standby con- 

. The third subject dealt with by the Commit- 
tee was that of the National Science Founda- 
tion. At the time of the report, it was thought 
that a compromise bill acceptable to both the 
Congress and the Administration would be in- 
troduced soon. 

The recent report of the President's Scien- 
tific Research Board was reviewed and was 
compared with a similar study made by Rosa 
after World War I.* The fact that many of the 
recommendations regarding the scientific pro- 
gram of the Government made in the Rosa re- 
port are similar to those in the current report 
was cited by the Committee in support of their 
recommendation that the Academy give serious 
and sustained attention to legislation relating 
to science. 

A. T. McPherson, Chairman 



Dr. Waldo L. Schmitt, President of the 
Washington Academy of Sciences in 1947, ap- 
pointed a committee on junior academies of 
science consisting of Austin H. Clark, Chair- 
man, E. H. Walker, Vice Chairman, and Frank 
Thone, and requested it to study the subject 
and make recommendations to the Washington 
Academy of Sciences. The following is the re- 
port of the vice chairman based upon the results 
of the study as given in his talk at the 50th 

* Expenditures and revenues of the Federal Gov- 
ernment, by Edward B. Rosa. Publication No. 
1518. Reprinted from "Taxation and Public Ex- 
penditures," Vol. 95 of the Annals of the Ameri- 
can Academy of Political and Social Science, 
Philadelphia, May 1921. 

June 15, 1948 

proceedings: the academy 


annual meeting of the Washington Academy on 
January 15, 1948. 

The growing interest in the early recognition 
and encouragement of potential scientists has 
found expression in the development of organi- 
zations working with and for these young scien- 
tists. Their activities are centered largely on 
the approximately 15,000 science clubs in the 
secondary schools throughout the country. 
Assisting in the coordinating of these clubs is 
Science Clubs of America, administered by 
Science Service, Inc., Washington, D. C. An 
important activity of this coordinating organi- 
zation is the conducting of the annual Science 
Talent Search for the Westinghouse Science 
Scholarships. Scholarships totaling $11,000 are 
distributed among the 40 winners on the basis 
of a nation-wide examination. In 11 states 
additional awards are granted to local winners 
in this same examination, this usually under 
the auspices of the state junior or senior acade- 
mies of science. 

Junior academies of science exist in more 
than 30 states. The first one was formed in Il- 
linois in 1919. These organizations provide to 
the young scientists and their sponsors, mostly 
organized into school science clubs, (1) valuable 
training through cooperative endeavors; (2) 
association with other young scientists and 
with adult scientists; (3) incentives to the stu- 
dents through exhibits, fairs, and congresses, to 
engage in creative activities; (4) incentives and 
assistance to the adult sponsors; (5) insight 
into senior scientific organizations and their ac- 
tivities; and (6) opportunities to take an active 
part in adult scientific projects. 

Membership in these junior academies usu- 
ally consists of the science clubs as chapters and 
their members as individuals. They are vari- 
ously financed, the funds coming from dues 
paid by the chapters or by the members, or 
both, and subsidies by senior academies, 
teacher organizations, sponsors, friends, com- 
mercial organizations, etc. Sums run from 
about $20 to $400 annually. Additional services 
and facilities are often provided by associated 
and sponsoring organizations. 

The Junior academies of science are usually 
governed by committees of the senior acade- 
mies and of the school sponsors of the clubs. 
They function best where there is an experi- 
enced and active adult leader on both commit- 
tees or at least actively supported by them. 

These junior academies, besides assisting in 
the conduct of the science clubs' activities, 
provide stimulating and coordinating publica- 
tions, sectional meetings, an annual state con- 
gress, usually in connection with the annual 
state academy meeting, with exhibits, fairs, 
trips, talks, etc., and awards of various kinds. 
They often enlist the aid of juniors in various 
senior scientific projects where such is appro- 

The organization of a Junior Washington 
Academy of Science is favored by local indi- 
viduals and organizations engaged in junior 
scientific work. The head of the science work in 
the public schools and the teacher-sponsors of 
the science clubs believe it would be a spur to 
their efforts and could provide valuable facili- 
ties they are otherwise unable to obtain. Sci- 
ence Clubs of America would be glad for such 
an Academy to take over its responsibilities in 
the annual science fair, an activity which it has 
previously assumed in the absence of other 
suitable organizations to do so. It would be glad 
to assist in the selection of the best science 
talent in this area for awards in addition to the 
national Westinghouse Science Scholarships. 

This committee's investigation of the local 
organizations and individuals who might be in- 
terested in this project is incomplete. More 
work should be done in this field before a final 
plan of organization is prepared. There should 
be careful investigation of the activities of the 
Virginia Junior Academy of Science, centered 
in Charlottesville, in order to coordinate the 
considered activities of a Junior Washington 
Academy of Sciences with those of this nearby 
organizatioD working within the greater Wash- 
ington area. We have had a preliminary meet- 
ing with officers of the Maryland Academy of 
Sciences in Baltimore and find that our activi- 
ties would supplement theirs in the adjacent 
portion of Maryland, with which we might be 
concerned, and that cooperation with that or- 
ganization would be highly desirable and wel- 
comed. Their junior academy is temporarily 
suspended for lack of a director, but much 
junior work is being carried on along with their 
program of interpretation of science. The dif- 
ferent emphases and perspectives of the Mary- 
land and of the Washington Academies of Sci- 
ence would find common ground through this 
proposed Junior Washington Academy of Sci- 



We, therefore, recommend that the Board in 
control of the Washington Academy of Sciences 
continue this committee with new or aug- 
mented personnel and instruct it to report at 
the April meeting of the Academy definite and 
detailed plans for organizing a Junior Washing- 
ton Academy of Sciences with proposals for 
implementing this plan. 

Egbert H. Walker, Vice Chairman 

report of special committee to consider 

various matters pertaining to the 

Journal and its improvement 

The report of this committee was included in 
the published Proceedings of the Academy, 
414th Meeting of the Board of Managers: This 
Journal 38 (2): 79-80. Feb. 15, 1948. 

Submitted by C. Lewis Gazin, Secretary 


The 417th meeting of the Board of Mana- 
gers, held in the Cosmos Club, April 12, 1948, 
was called to order at 8:05 p.m. by the Presi- 
dent, Dr. F. D. Rossini. Others present were: 
H. S. Rappleye, N. R. Smith, H. A. Rehder, 
W. W. Diehl, W. Ramberg, T. D. Stewart, 
C. E. White, A. Wetmore, W. A. Dayton, 
C. A. Betts, M. A. Mason, L. A. Rogers, 
C. L. Gazin, and, by invitation, H. E. Mc- 
Comb, R. J. Seeger,- A. Stone, J. E. Graf, 
and F. H. H. Roberts, Jr. 

The following appointments were announced 
by the President: Committee on the Index of 
the Journal: W. N. Fenton, Chairman, J. I. 
Hoffman, M. C. Merrill, H. A. Rehder, 
Paul H. Oehser (consultant). Committee on 
Functions and Policies of the Academy: E. C. 
Crittenden, Chairman, A. H. Clark, W. A. 
Dayton, M. A. Mason, L. W. Parr, F. B. 


The Chairman of the Meetings Committee 
announced that Prof. Philipp Frank, of Har- 
vard University, would speak at the April 
meeting of the Academy. 

The Chairman of the Special Committee to 
consider certain revisions of the Bylaws of the 
Academy and Standing Rules of the Board of 
Managers, Dr. Walter Ramberg, read the 
following report: 

The Committee has considered means of revis- 
ing the Standing Rules of the Board of Managers 

in order to provide greater continuity in the 
Standing Committees of the Board. The Commit- 
tee recommends the following revisions for this 

Standing Rules 

Section 3, third sentence, replace by: "The Com- 
mittee on Membership, Meetings, Grants-in- 
Aid for Research, and each of the Subcommittees 
of the Committee on Awards for Scientific 
Achievement shall include, if possible, at least two 
members reappointed from the preceding year. 
The Committee on Monographs shall have six 
members, each appointed for 3 years at the rate 
of two each year. At least three members of the 
Committee on Monographs shall be past editors 
of the Journal, if possible." 

In addition the Committee recommends the 
following revision in the Standing Rules in order 
to describe more accurately the duties of the Com- 
mittee on Awards for Scientific Achievement: 

New Section 8, second sentence, replace by: "A 
candidate must be a member of the Washington 
Academy of Sciences or a resident member of one 
of the affiliated societies, and shall not have 
passed the 40th anniversary of the date of his 
birth by the end of the calendar year for which 
the award is made; recommendations by the 
Committee must reach the Board of Managers 
not later than the meeting preceding the Annual 
Meeting of the Academy in January." 

The Chairman of the Committee on Science 
Legislation, John E. Graf, reported that 
Senate Bill 526 had been vetoed by President 
Truman but that Senator Smith of New Jersey 
was sponsoring S3285, introduced March 25, 
which would overcome some of the objections 
raised to the previous bill. 

The president announced that the results of 
the study to be made by the Special Commit- 
tee appointed to consider the policy and func- 
tions of the Academy were to be available by 
the end of the calendar year and to be pub- 
lished in the Journal as a matter of permanent 
record and in order to bring them to the atten- 
tion of the membership of the Academy. 

Changes in the Standing Rules of the Board 
of Managers introduced at the 416th meeting 
were approved as follows: 

Section 2 

After c. iv., insert: 

"v. Committee on Awards for Scientific 

"vi. Committee on Grants-in-aid for Re- 

Section 3 

Replace first sentence by: "There shall be six 
Standing Committees, as follows: Executive 

June 15, 1948 



Committee, Committee on Meetings, Commit- 
tee on Membership, Committee on Mono- 
graphs, Committee on Awards for Scientific 
Achievement, Committee on Grants-in-Aid for 

Section 4, first sentence: 

In place of "or acting President," insert "Presi- 
dent-Elect" (providing the Academy approves 
of the new office of President-Elect). 

Following Section 6, insert: 

7. The Committee on Monographs shall re- 
ceive for review and recommendation regard- 
ing publication such monographs as may be 
submitted to the Academy, preference being 
given to members of the Academy. 

8. It shall be the duty of the Committee on 
Awards for Scientific Achievement of 
recommending one candidate each for the Bio- 
logical Sciences, the Engineering Sciences, and 
the Physical Sciences. A candidate shall not 
have passed the 40th anniversary of the date of 
his birth by the end of the calendar year for 
which the Award is made; recommendations 

by the Committee must reach the Board of 
Managers not later than the meeting immedi- 
ately preceding the annual meeting of the 
Academy in January. Each recommendation 
to the Board must be accompanied by a written 
supporting statement giving the necessary in- 
formation concerning the candidate, together 
with a citation covering not over 80 spaces, as 
"in recognition of his distinguished service in 

(80 spaces) ." 

9. The Committee on Grants-in-Aid for 
Research shall review applications for grants- 
in-aid from such funds as may be at the dis- 
posal of the Board of Managers for this purpose. 


present Section 7 as 
present Section 8 as 
present Section 9 as 
present Section 10 a; 
present Section 11 a 
present Section 12 ai 

new Section 10 
new Section 11 
new Section 12 
5 new Section 13 
5 new Section 14 
5 new Section 15. 

The meeting was adjourned at 9:35 p.m. 
C Lewis Gazin, Secretary. 


George Rogers Mansfield, retired geolo- 
gist of the U. S. Geological Survey, member of 
the Geological Society of Washington, and its 
39th President in 1930, died at the home of one 
of his daughters, Mrs. John W. Carroll, at 
Swarthmore, Pa., on Sunday, July 11, 1947. He 
was born August 30, 1875, in Gloucester, Mass. 
At Amherst College he received the B.S. degree 
and membership in Phi Beta Kappa in 1897, 
and the M.A. degree in 1901. From 1897 to 
1903 he was science teacher at Central High 
School in Cleveland, Ohio. At Harvard Uni- 
versity he received the Ph.D. degree in 1906 
and was instructor in geology from 1906 to 
1909. At Northwestern University he was as- 
sistant professor in geology from 1909 to 1912. 
He worked during the summers of 1910 to 1912 
for the U. S. Geological Survey in Idaho. In 
1913 he joined the permanent staff as geologist 
and was placed in charge of phosphate investi- 
gations, which he carried on for many years. In 
1922 he was made chief of the Section of Non- 
metalliferous Deposits, in 1921 editor of geolog- 
ic maps, and in 1927 chief of the Section of 
Areal and Nonmetalliferous Geology. He re- 
tired in 1943. 

Dr. Mansfield's contributions to geology 
over a period of 36 years (1906-42), include 109 
titles, of which Professional Paper 152: Geog- 

raphy, geology, and mineral resources of a part of 
southeastern Idaho, published in 1927, may be 
cited as an example of his outstanding com- 
petence in investigation, description, discussion 
and solution of the areal, physiographic, strati- 
graphic, structural, and economic problems en- 
countered in the detailed study of an area (in 
this case of some 2,200 square miles). Other 
publications of the Survey of which he was 
either author or joint author described the 
phosphates of Florida, the greensands of New 
Jersey, the nitrates of California, Texas, Idaho, 
and Oregon, and the potash deposits of Texas 
and New Mexico. His other contributions ap- 
peared in many publications, mainly those of 
the Geological Society of America, American 
Association of Geographers, the American 
Journal of Science (of which he was associate 
editor from 1938 to 1945), the Journal of Ge- 
ology, Washington Academy of Sciences, 
Industrial and Engineering Chemistry, Eco- 
nomic Geology, Science, Proceedings of Section 
E of the American Association for the Advance- 
ment of Science. 

He was a member of the National Research 
Council 1925 to 1934 and chairman of the com- 
mittee on tectonics from 1924 to 1934, a fellow 
of the Geological Society of America, and a 
member of the American Institute of Mining 


and Metallurgical Engineers, the Society of versity, Dr. James S. Mansfield, of Cambridge, 
Economic Geologists, the American Geophysi- Mass., and Robert H. Mansfield, of Caracas, 
cal Union, the Washington Academy of Sci- Venezuela; two daughters, Mrs. George W. 
ences (vice president 1931), and the Cosmos Patterson, of Morton, Pa., and Mrs. John W. 
Club. Carroll, of Swarthmore, Pa.; and twelve grand- 
He is survived by his wife, Adelaide Clafin children. 
Mansfield; three sons, Harvey C. Mansfield, a Ralph W. Richards 
professor of political science at Ohio State Uni- 

Officers of the Washington Academy of Sciences 

President Frederick D. Rossini, National Bureau of Standards 

Secretary C. Lewis Gazin, U. S. National Museum 

Treasurer Howard S. Rappleye, Coast and Geodetic Survey 

Archivist Nathan R. Smith, Plant Industry Station 

Custodian and Subscription Manager of Publications 

.Harald A. Rehder, U. S. National Museum 

Vice-Presidents Representing the Affiliated Societies: 

Philosophical Society of Washington Walter Ramberg 

Anthropological Society of Washington T. Dale Stewart 

Biological Society of Washington John W. Aldrich 

Chemical Society of Washington Charles E. White 

Entomological Society of Washington C. F. W. Muesebeck 

National Geographic Society Alexander Wetmore 

Geological Society of Washington William W. Rtjbey 

Medical Society of the District of Columbia Frederick O. Coe 

Columbia Historical Society Gilbert Grosvenor 

Botanical Society of Washington Ronald Bamford 

Washington Section, Society of American Foresters William A. Dayton 

Washington Society of Engineers Clifford A. Betts 

Washington Section, American Institute of Electrical Engineers 

Francis B. Silsbee 

Washington Section, American Society of Mechanical Engineers 

Martin A. Mason 

Helminthological Society of Washington Aurel O. Foster 

Washington Branch, Society of American Bacteriologists Lore A. Rogers 

Washington Post, Society of American Military Engineers . Clement L. Garner 

Washington Section, Institute of Radio Engineers Herbert Grove Dorsey 

Washington Section, American Society of Civil Engineers Owen B. French 

Elected Members of the Board of Managers: 

To January 1949 Max A. McCall, Waldo L. Schmitt 

To January 1950 F. G. Brickwedde, William W. Diehl 

To January 1951 Francis M. Defandorf, William N. Fenton 

Board of Managers All the above officers plus the Senior Editor 

Board of Editors and Associate Editors [See front cover] 

Executive Committee Frederick D. Rossini (chairman), Walter Ramberg, 

Waldo L. Schmitt, Howard S. Rappleye, C. Lewis Gazin 

Committee on Membership 

Harold E. McComb (chairman), Lewis W. Butz, C. Wythe Cooke, William 

W. Diehl, Lloyd D. Felton, Regina Flannery, George G. Manov 

Committee on Meetings Raymond J. Seeger (chairman), 

Frank P. Cullinan, Fred L. Mohler, Francis O. Rice, Frank Thonb 

Committee on Monographs: 

To January 1949 Lewis V. Judson (chairman), Edward A. Chapin 

To January 1950 Roland W. Brown, Harald A. Rehder 

To January 1951 William N. Fenton, Emmett W. Price 

Committee on Awards for Scientific Achievement (Karl F. Herzfeld, general chairman): 

For the Biological Sciences 

C. F. W. Muesebeck (chairman), Harry S. Bernton, Chester W. Emmons, 
Elmer Higgins, Mario Mollari, Gotthold Steiner, L. Edwin Yocum 

For the Engineering Sciences 

Harry Diamond (chairman), Lloyd V. Berkner, Robert C. Duncan, 
Herbert N. Eaton, Arno C. Fieldner, Frank B. Scheetz, W. D. Sutcliffe 

For the Physical Sciences 

Karl F. Herzfeld (chairman), Nathan L. Drake, Lloyd D. Felton, 
Herbert Insley, William J. Rooney, Robert Simha, Michael X. Sullivan 

Committee on Grants-in-aid for Research 

. .F. H. H. Roberts, Jr. (chairman), Anna E. Jenkins, J. Leon Shereshevsky 

Representative on Council of A. A. A. S Frank Thone 

Committee of Auditors 

William G. Brombacher (chairman), Harold F. Stimson, Herbert L. Haller 

Committee of Tellers , 

...John W. McBurney (chairman), Roger G. Bates, William A. Wildhack 



Physics. — Mass spectra of hydrocarbons. Fred L. Mohler 193 

Botany. — New species of Achaetogeron (Compositae) from Mexico. 

Esther L. Larsen 199 

Entomology. — Notes and descriptions of Nearctic Hydroptilidae 

(Trichoptera). Herbert H. Ross 201 

Zoology. — An analysis of specific homonyms in zoological nomen- 
clature. Richard E. Blackwelder 206 

Proceedings: The Academy 213 

Obituary: George Rogers Mansfield 223 

This Journal is Indexed in the International Index to Periodicals 

Vol. 38 

July 15, 1948 

No. 7 





James I. Hoffman 


Alan Stone 


Frank C. Kracek 



Lawrence A. Wood 


J. P. E. Morrison 


Elbert L. Little, Jr. 


Richard E. Blackwelder 


James S. Williams 


Waldo R. Wedel 


Irl C. Schoonover 




450 Ahnaip St. 

at Menasha, Wisconsin 

Entered as second class matter under the Act of August 24, 1912, at Menasha, Wis. 

Acceptance for mailing at a speoial rate of postage provided for in the Act of February 28, 1925. 

Authorised January 21, 1933. 

Journal of the Washington Academy of Sciences 

This Journal, the official organ of the Washington Academy of Sciences, publishes: 
(1) Short original papers, written or communicated by members of the Academy; (2) 
proceedings and programs of meetings of the Academy and affiliated societies; (3) 
notes of events connected with the scientific life of Washington. The Journal is issued 
monthly, on the fifteenth of each month. Volumes correspond to calendar years. 

Manuscripts may be sent to any member of the Board of Editors. It is urgently re- 
quested that contributors consult the latest numbers of the Journal and conform their 
manuscripts to the usage found there as regards arrangement of title, subheads, syn- 
onymies, footnotes, tables, bibliography, legends for illustrations, and other matter. 
Manuscripts should be typewritten, double-spaced, on good paper. Footnotes should 
be numbered serially in pencil and submitted on a separate sheet. The editors do not 
assume responsibility for the ideas expressed by the author, nor can they undertake to 
correct other than obvious minor errors. 

Illustrations in excess of the equivalent (in cost) of one full-page halftone are to 
be paid for by the author. 

Proof. — In order to facilitate prompt publication one proof will generally be sent 
to authors in or near Washington. It is urged that manuscript be submitted in final 
form; the editors will exercise due care in seeing that copy is followed. 

Unusual cost of foreign, mathematical, and tabular material, as well as alterations 
made in the proof by the author, may be charged to the author. 

Author's Reprints. — Reprints will be furnished in accordance with the following 
schedule of prices (approximate) : 

Copies 4 pp. 8 pp. 12 pp. 16 pp. 20 pp. Covers 

50 $2.40 $3.90 $ 6.25 $ 6.55 $ 8.70 $2.40 

100 3.00 4.80 7.70 8.10 10.50 3.30 

150 3.60 5.70 9.15 9.70 12.30 4.20 

200 4.20 6.60 10.60 11.25 14.10 5.10 

250 4.80 7.50 12.00 12.80 15.90 6.00 

Subscriptions or requests for the purchase of back numbers or volumes of the Jour- 
nal or the Proceedings should be sent to Harald A. Rehder, Custodian and Sub- 
scription Manager of Publications, U. S. National Museum, Washington 25, D. C. 

Subscription Rates for the Journal. — Per year $7 .50 

Price of back numbers and volumes: Per Vol. Per Number 

Vol. 1 to vol. 10, incl. — not available.* — — 

Vol. 11 to vol. 15, incl. (21 numbers per vol.) $10.00 $0.70 

Vol. 16 to vol. 22, incl. (21 numbers per vol.) 8.00 0.60 

Vol. 23 to current vol. (12 numbers per vol.) 7 .50 .90 

* Limited number of complete sets of the Journal (vol. 1 to vol. 37, incl.) available 
for sale to libraries at $318.50. 

Proceedings, vols. 1-13 (1899-1911) complete $25.00 

Single volumes, unbound 2 .00 

Single numbers 25 

Missing Numbers will be replaced without charge provided that claim is made to the 
Treasurer within 30 days after date of following issue. 

Remittances should be made payable to "Washington Academy of Sciences" and 
addressed to the Treasurer, H. S. Rappleye, U. S. Coast and Geodetic Survey, Wash- 
ington 25, D. C. 

Exchanges. — The Academy does not exchange its publications for those of other 




Vol. 38 

July 15, 1948 

No. 7 

ARCHEOLOGY. — Florida archeology and recent ecological changes. 1 John M. 
Goggin, Peabody Museum of Natural History, Yale University. (Com- 
municated by W. R. Wedel.) 

American archeology in the past few 
years has turned to several of the natural 
sciences for aid in dating recent sites and 
for a fuller understanding of the natural 
environment in which primitive man played 
his role. In the Southwest, for example, 
archeologists have been able to obtain ca- 
lendric or absolute dates from the study of 
tree rings (Glock, 1937; Haury, 1935). 
Pollen analysis, of considerable value in 
European prehistory, also shows promise of 
utility in the New World (Sears, 1937). 
Other methods of dating and interpretation 
are based upon work done by the geologist 
and conchologist (Baker, 1937; Richards, 

The best results are now being obtained 
by cooperative research on the part of both 
archeologists and natural scientists. Two 
such studies can be briefly summarized: 
One in Oregon gave a relative date for ar- 
cheological material sealed in a cave deposit 
under volcanic ash. The same ash layer was 
found in peat deposits. By pollen analysis 
the history of the peat deposit was worked 
out, and the relative date of this ash layer 
was determined, which in turn gave an 

1 Data used in this paper were gathered as part 
of the Yale Caribbean Program directed by Dr. 
Cornelius Osgood. T*he 1947 summer's work was 
done under a fellowship from the Social Science 
Research Council. Many persons have been very 
helpful in offering me unpublished data, especially 
Irving Rouse, Vera M. Masius, and John W. 
Griffin. Others have generously read and criti- 
cized the manuscript. For this special thanks are 
due Frederick Johnson, R. S. Peabody Founda- 
tion, Phillips Academy, Andover, Mass.; Irving 
Rouse, Department of Anthropology, Yale Uni- 
versity; Richard F. Flint, Department of Ge- 
ology, Yale University; and Martin Burke nroad 
Newport, North Carolina. Received March 22, 

upper limit relative date for the archeo- 
logical material in the cave. 2 

Another recent example of a combined 
project of this type is the Boylston Street 
fishweir. In the course of a construction 
excavation in Boston the remains of a fish- 
weir were found many feet below the pres- 
ent surface of the land. The indications are 
that man lived in the Charles River estuary 
"when the level of the sea in relation to 
land, was about fifteen feet eight inches 
lower than it is at the present time" (John- 
son et al., 1942). A more recent joint prob- 
lem of archeology and botany is the Grassy 
Island site (Johnson and Raup, 1947). 3 

Approaches like these were originally 
developed in the Old World where the usual 
time span involved was much greater. 4 Al- 
though American archeologists had been 
aware of the European results, it was 
thought until recently that the length of 
time in which the greatest developments in 
American cultures took place was too short 
to have resulted in ecological changes of any 
importance. It is now realized that while 
the changes that have taken place in the 
New World are not so sharp as those in 
some parts of Europe, such as in Scandi- 
navia, nevertheless they are present and 
can be detected by more refined analysis. 

2 Symposium on Early Man in Oregon: Cress- 
man, 1946; Hansen, 1946; Allison, 1946. Han- 
sen, 1947. 

3 As a point of historical interest it can be noted 
that E. S. Morse was perhaps one of the first to 
point out a series of ecological changes in vegeta- 
tion, species of animals, and erosional factors in a 
single Maine shell heap (1868; 1925: 430). 

4 A complete summary of the methods and re- 
sults of archeological and geological dating will be 
found in Zeuner, 1946. 


u c 



The twofold problem facing the archeolo- 
gist can be briefly summarized as follows: 
The first phase will necessitate the prepara- 
tion of a detailed ecological history of late 
Pleistocene and Recent times covering the 
climate, the composition and changes in 
flora and fauna, and the changes in land 
form, especially the coast line. These are 
factors directly affecting primitive man on 
a simple subsistence level. The second 
problem, of special interest to the archeolo- 
gist, is whether any of these changes can be 
used 'for either absolute or comparative 
dating. A hope for absolute dating possibly 
lies in the analysis of some constant process 
such as tree rings or the deposition of sedi- 
ments like varves or peats, and less likely 
in sea-level changes. But sea-level changes, 
as well as ecological changes, do offer hope 
of comparative correlations with other 
regions where similar conditions can be 

The analysis of climatic fluctuations with 
the attendant floral and faunal changes also 
offers much hope for comparative correla- 
tions. This should be especially helpful in 
Florida where a relatively sharp temperate 
versus tropical physiographic line can now 
be drawn. Minor climatic changes, only 
scantily reflected in either more temperate 
or tropical regions, probably resulted in 
appreciable movements of the border line 
in Florida. 5 This meant that abrupt changes 
often took place in Florida, which necessi- 
tated adjustment of the biota to the new 

In the course of recent archeological work 
in Florida the writer has noticed a number 
of situations that may offer valuable infor- 
mation if they can be considered in detail 
by competent specialists. It is the object of 
this paper to bring these situations to the 
attention of the respective specialists in the 
hope that they may be stimulated to follow 
up some of the problems, or to instill an 

5 It is probably no accident that the bulk of the 
comparable data for ecological changes in Eastern 
United States, mentioned in later sections of this 
paper, comes from Maine. This too is a physio- 
graphic boundary area and climatic variances 
would make appreciable changes here as in 
Florida. An area in the middle of a large physio- 
graphic region, Virginia for example, would not be 
affected by variances which would result in pro- 
nounced changes in Florida or Maine. 

awareness of the problems in event that 
similar situations arise in routine biological 
or geological research. Some of the following 
situations have been directly observed in 
the field and others are taken from existing 
literature. In most cases there is not enough 
information to formulate concrete conclu- 
sions. They must be thought of as inter- 
esting leads offering promise with future 
detailed work. 

No attempt has been made to gather 
similar data from purely natural-history 
sources or to evaluate the material pre- 
sented here in terms of other ecological 
data. To repeat, it is the purpose of this 
paper to present evidence of ecological 
changes that can be related to human ac- 
tivity in this region for dating purposes. An 
explanation of the various archeological 
cultures referred to in this article will be 
found in a recent paper (Goggin, 1947). A 
chronological chart from that paper is re- 
produced here (Fig. 1). 

Changes in sea level. — Careful measure- 
ments in recent years have brought forth 
data on the rise of sea level relative to a 
stable coast line (Flint, 1947: 426-428). In 
south Florida this rise has been demon- 
strated by Davis (1940:402-405; 1946: 180- 
181) in his study of the mangrove swamp. 

It is of interest therefore to be able to 
correlate this rise in sea level with sites oc- 
cupied by man. Numerous midden sites in 
southern Florida, particularly those on the 
lower east coast, in the Ten Thousand Is- 
lands, and in the Cape Sable region, which 
are now small islands in the mangroves, 
were probably on dry land when first oc- 
cupied. A good example is one located in 
the present village of Surfside north of 
Miami Beach. This site consists of a black 
dirt and shell midden with an adjacent 
burial mound made of beach sand. The mid- 
den site was first occupied in Glades II 
times, but the burial mound is believed to 
date from Glades III times (Fig. 1). Pre- 
vious to the filling operations the site was 
a small dry, hammock-covered island in 
the mangrove swamp. 

Excavation in the burial mound has re- 
vealed the lowermost group of human bones 
to be completely below low tide level. No 
mangrove peat underlies the mound. It is 

July 15, 1948 goggin: Florida archeology and ecological changes 


very unlikely that the burials were made 
under water and then covered with sand; 
therefore the sea level must have risen since 
the burials were deposited. This postulation 
is substantiated by trenches in the midden 
part of the site, which show that deposit to 
be resting on clean beach sand, not man- 
grove peat. 

A reconstruction of the history of the 
site would be as follows: The first occupa- 
tion here was on the dry inner shore of the 
beach ridge. Subsequently the rise of sea 
level inundated the low shoreline and iso- 
lated the site, which by that time had been 
built up enough to be above the rising sea. 
Thus the site became an island which in 
time was surrounded by a mangrove swamp. 

Changes in ground-water level. — In parts 
of southern Florida there has been an ap- 
preciable demonstrable rise in ground- water 
level, which is in some places associated 
with a deposition of sediments. 

South Indian Field is a sand midden site 

in the valley of the St. Johns River near its 
head, west of Malabar, Brevard County. 
This little cabbage-palm hammock is a 
mound composed of sand and refuse about 
3 feet higher than the surrounding sawgrass 
prairie (Rouse, MS.). Previous to modern 
drainage water normally surrounded the 
site, and often only the constricted highest 
portions were above water. 

Excavations in the site (Masius, MS.) 
indicate that the first occupation began here 
on a level approximately the same as the 
present surface of the prairie. As refuse ac- 
cumulated the site was built to its present 
height. In terms of the archeological picture 
this site was fairly early, its initial occupa- 
tion in the Orange Period being about 700 
A. D. or earlier (Fig. 1). 

An interesting problem is posed here 
(Rouse, MS.). In the light of recent con- 
ditions, the earliest level of the site could 
not have been a suitable camp site for the 
Indians, for it was never dry enough. Even 




COAST ( 1 ) 

COAST (2) 









































n ! 








I (?) 





















Fig. 1. — Archeological areas and periods in Florida. The term "Archaic Horizon," as used 
in the text, refers to the Orange, Tick Island, and Nonceramic periods. 



VOL. 38, NO. 7 

a short seasonal dry spell or several years of 
drought would not have given a sufficient 
period of time for the beginnings of the 
midden, as the deposit shows no signs of a 
rapid accumulation which might have en- 
abled its builders to get above the present 
water level. Therefore, we have to postu- 
late a different initial ecological condition, 
one in which the prairie was completely dry. 
Since that time the water table has risen 
and the surrounding area become a marsh. 

As corroborative evidence for the dry- 
ness of the locality Rouse (MS.) has noted 
that the present surface material of the 
prairie is sand, but until recently there was 
a thin layer of muck or peat, which has 
burned off since drainage. There is no muck 
under the refuse deposit; so such material 
must have formed since the development of 
the site. In addition there are traces of a 
large number of pits which were dug around 
the site during the early period of occupa- 
tion. They seem admirable wells, but other- 
wise their presence is difficult to account 
for. Under present conditions (that is, be- 
fore drainage) wells are not needed, but if 
the prairie was once dry they would have 
been very necessary. 

Several possible theories to account for 
the apparent history of this locality can be 
given, but field study is necessary to deter- 
mine the true cause. A rise in sea level 
would result in a higher ground- water table, 
but this could only account for part of the 
rise. Increased precipitation would certainly 
be important, and another possible factor 
would be the damming up of the St. Johns 
Valley farther down the river, by the de- 
velopment of extensive peat beds. 

Archeological sites in the northern Ever- 
glades, on the south shore of Lake Okeecho- 
bee, are found in conditions indicating con- 
siderable rise in ground water level along 
with sedimentation. The large midden site 
at Chosen near Belle Glade has a foot of 
muck over its edges, while the bottom was 
two feet below the 1934 ground surface 
(Stirling, 1935: 374; Willey, MS.). A con- 
sideration of local muck subsidence prob- 
ably would increase this depth figure. The 
midden must have been begun at some 
period of low water when the Everglades 
were dry or nearly so, possibly at one of the 

low water stages indicated by Dachnowski- 
Stokes (1930). 

At South Bay, to the west of Belle Glade, 
even deeper archeological sites have been 
reported. 6 These are small midden deposits 
with animal bones and pottery which lay 
about three feet below the present surface 
of the muck. A number of these have been 
noted in the sides of drainage ditches in this 
region. All appear to be on the same strati- 
graphic horizon. The reported two-foot 
subsidence in this region gives an original 
depth of five feet for these small sites. Like 
the Belle Glade site, they could not have 
been formed under anything similar to the 
modern predrainage conditions, but must 
have been begun at a relatively dry period. 

Changes in local ecology. — There is some 
archeological evidence that coastal lagoons 
such as the Indian River and Halifax River 
were once much fresher than in recent years. 

The Florida Archaic peoples, using fiber- 
tempered pottery, lived mainly along the 
St. Johns River, where their large shell 
middens indicate the use of almost all the 
available forms of Mollusca and animals for 
food. Evidence of their residence along the 
Atlantic coast is scanty but two sites have 
been examined and one more is reported. 
These two sites, one near Malabar on the 
beach ridge (Rouse, MS.) and the other on 
the west shore of Halifax River north of 
Ormond (Blatchley, 1902), 7 are unique in 
that they are mainly composed of coquina 
shells (Donax sp.) with occasional moon 
shells. Other shells are uncommon but in- 
clude fresh-water forms and occasional oy- 
sters. The latter site is an unusually large 
deposit, being over 100 yards in length and 
up to 12 feet in depth, while the former is 
much shallower. 

The composition of these sites is unusual 
because other archeological sites on this 
coast are predominantly formed of oyster 
shells, with only occasional small pockets 
of coquina shells. Moreover, the oyster 

6 Personal communication from L. M. Hardy of 
South Bay, a University of Florida student, July 
10, 1947. Unfortunately, high water made it im- 
possible to examine these sites in the 1947 sum- 
mer season. 

7 Excavations here in the summer of 1947 by 
the Florida Park Service will give us much more 
data when the results are studied. 

July 15, 1948 goggin: Florida archeology and ecological changes 


shell heaps are all of the St. Johns I and II 
periods dating from a. later time (Fig. 1). 
Therefore, the difference in composition of 
these sites has temporal significance. 

It seems probable that if oysters were 
available in quantity they would have been 
eaten by the Archaic peoples. They cer- 
tainly showed little selectivity for foods on 
the St. Johns, and the occasional oyster 
shell in the coastal middens indicates that 
they were used when available. Under pres- 
ent ecological conditions (before dredging) 
and for some time previous, judging from 
the St. Johns 1 and II shell heaps, oysters 
were abundant in the coastal lagoons. By 
inference from their absence in the early 
sites, it seems probable that at the time 
these middens were deposited oysters were 
not available in quantity. 

A successful oyster habitat is one that 
has the proper concentration of brackish 
water; therefore oysters usually grow in 
large bays and estuaries where fresh water 
constantly mingles with salt water. If this 
proper habitat was not available in Archaic 
times then the water must have been either 
too fresh or too salty. The possibility of its 
being too salty is not so probable in view of 
the lower sea level, unless there was a period 
of drought which reduced the supply of 
fresh water from the interior. It is more 
probable that the level of the sea was 
enough lower (perhaps no more than a 
couple of feet) to allow fresh water or only 
slightly brackish water to stand in these 
shallow lagoons. A similar situation could 
be the result of greater runoff, from in- 
creased precipitation, in the feeding streams. 
However, the shallowness of these coastal 
lagoons favors the former theory. Rein- 
forcing evidence for this is the fact that a 
slightly lower sea level would be accom- 
panied by a steeper gradient in the streams 
emptying into the lagoons, which would 
probably have resulted in a greater dis- 
charge of fresh water. In this flat country 
with the short streams such a change in 
gradient would be of importance. 

Another possible explanation for the fresh- 
ness of lagoon water takes into considera- 
tion the alternate opening and closing of 
inlets along the coast. This is an observed 
situation noted from the late eighteenth 

century (Romans, 1775: 287) to the present. 
However, inlet closing would produce 
purely local phenomena while we are deal- 
ing with areas as far apart as Ormond and 
Malabar, but which appear to share the 
same ecological situations. 

Another possibility, which seems least 
probable to the writer, is that there was 
little or no change in depth and salinity of 
the coastal lagoons and that the absence of 
oysters must be due to other causes. In 
event of such a situation the possibility of 
some catastrophe killing the oysters cannot 
be ignored. Cold waves, salinity changes, 
and micro-organisms are known to have 
caused much damage among fish, although 
the damage to invertebrates is not well 
known (Gunter, 1947). Such disasters would 
be much more severe in the coastal lagoons 
than along the open Atlantic. Thus Mol- 
lusca such as the coquina would be less 

It has been suggested (personal com- 
munication, Martin Burkenroad, January 
1948) that considerable time might have 
been necessary after the water had reached 
the proper salinity before oysters would 
have been numerous. This would have de- 
pended on the nature of the bottom; 
neither a very sandy nor a very muddy 
situation would have been conducive to a 
rapid growth of oyster beds. Under such 
conditions they would have developed 
slowly until the dead shells finally produced 
the proper base for an extensive growth. 

In summary we can reconstruct the situ- 
ation as follows : In late Archaic times — the 
Orange and Tick Island periods (Fig. 1) — 
there was little occupation of the north At- 
lantic coast of Florida. The few people 
there made shell middens of coquina shells 
gathered on sandy beaches. Because the 
coastal lagoons were too fresh, oysters were 
not common. In a few localities, perhaps at 
the mouths of inlets, some oysters were 
available and these were eaten when found. 
The freshness of the coastal lagoons may 
have been due to a number of factors, but 
the most probable is the lower sea level, 
with the possibility of increased precipita- 
tion being an additional factor. 

The problem of local ecological changes 
has been considered to some extent in 



studies of the famous Vero and Melbourne 
sites. Data on these finds are being reex- 
amined and evaluated by Irving Rouse 
(MS.) in terms of the newer ecological 
studies. It is probable that the history of 
these sites is closely tied to changes in the 
Indian River. 

In the northern Everglades around the 
south shore of Lake Okeechobee, peat pro- 
file studies have disclosed considerable local 
ecological changes (Dachnowski-Stokes, 
1930). Alternate low and high water stages 
are indicated by the sediments. This is 
mentioned here because of the evidence of 
deep archeological sites in these sediments 
(see the previous section). Unfortunately it 
is not clear if these variances represent local 
shoreline or other conditions or whether 
they are an expression of a broader climatic 
fluctuation. An examination of the biota of 
these beds might be fruitful. 

Changes in faunal composition. — The 
West Indies top shell, Livona pica Linnaeus, 
is an abundant shell in many middens along 
the lower east coast of Florida. It has been 
seen in almost every archeological site ex- 
amined between Key Largo and Sugarloaf 
Key in Monroe County and in numerous 
sites from Hillsborough Inlet north to 
Singers Island in Palm Beach County. This 
distribution coincides with the range of 
favorable habitat; that is, a rocky shore. 
These sites were occupied by the Indians 
in Glades II and III times, circa 1000-1500 
A.D. (Fig. 1). 

The presence of this shell is of consider- 
able interest because, with the exception of 
a couple of doubtful records, this mollusk 
is not now found in Florida (Clench and 
Abbott, 1943: 9). The present range of the 
species is the West Indies, and the nearest 
modern occurrence is their reported pres- 
ence on Dog Rocks, Cay Sal Bank, not far 
southeast of the Florida Keys. 

What is the significance of their presence? 
The large quantity of shell found in the 
sites and the close correlation of their oc- 
currence with their natural rock shore habi- 
tat suggests that the shells were locally 
collected, presumably for food, in the im- 
mediate vicinity of the sites. It is of interest 
to the archeologist whether their extinction 

in Florida can be attributed to man, to 
disease, or possibly to some environmental 
change which caused the animal to contract 
its range. If the last is the answer, was it 
purely a local manifestation or was it an 
expression of a broader change? 

Examples of Anodonta imbecillis Say 
(probably specimens of a local species 
closely related to A. imbecillis), a fresh- 
water shell, are described from the Ormond 
Archaic midden. Blatchley (1902: 72-74) 
cites Simpson to the effect that the present 
range of this mollusk is not south of North 

On the Florida Gulf Coast the Strombus 
gigas Linnaeus, is not now present, although 
there are records of dead shells or occasional 
specimens. 8 However, they are reported to 
occur in shell mounds at Cedar Keys (Cal- 
kins, 1878: 228; Ecker, 1878: 101). No shells 
of this species were seen in shell middens 
here in the course of a brief visit to this 
region, but a more intensive study may con- 
firm their presence. Most of these sites were 
occupied subsequent to 1000 A.D. 9 

Vertebrate remains have not generally 
been carefully identified in Florida arche- 
ology, but one excellent study does point 
out differences in faunal composition. At 
South Indian Field, west of Malabar, a 
wide variety of vertebrate bones were 

8 Worn dead specimens have been noted at 
Sanibel and Captiva Islands (Perry, 1940: 135). 
Strombus costatus rarely occurs in the vicinity of 
Tampa Bay (Simpson, 1887-89: 53). 

9 Evidence from other regions in the New 
World indicates that similar molluscan faunal 
changes have taken place in recent times. Oysters 
form a large part of Maine shell heaps, although 
they are now uncommon in that area. The scallop 
Pecten sp. is also found in these shell heaps al- 
though it now lives from Cape Cod to the south- 
ward, and the clam Mercenaria mercenaria is at 
present rare north of Cape Cod although it is 
abundant in Maine shell heaps (Morse, 1925: 430). 

In shellheaps on the Pearl Islands, west of the 
Isthmus of Darien, there were found three species 
of shells, Strombus peruvianus, Ostrea chilensis, 
and Solen (Tagelus) dombeyi, which do not have 
that northern range at the present time. To com- 
plicate things further there was found another 
species, Venus (Ventricola) rugosa, whose present 
habitat is the Antilles (Linne, 1929: 128-129). 

On the Rio Preto, a tributary of the Rio 
Grande da Conceicao in Brazil, Azara sp. shells 
were found (in middens) which are of a form not 
now occurring in the surrounding region (Koenigs- 
wald, 1905: 346; also see Serrano, 1946, Ihering, 
1903, and Krone, 1914). 

July 15, 1948 goggix: Florida archeology and ecological changes 


identified (Houck, MS.). Among them was 
an unworked jaw fragment of a beaver, 
Castor sp., found in the early Orange period 
(Fig. 1). This animal is not now found in 
the vicinity of the site, although beavers 
were reported in unspecified localities in 
northern Florida in Colonial times. 10 

Remains of the great auk have been 
found in a shell heap near Ormond (Hitch- 
cock, 1902; Blatchley, 1902). Details of 
the find are lacking, but the identification 
is apparently well confirmed. 11 This is the 
same site previously mentioned which was 
recently excavated by the Florida Park 

Changes in faunal size. — Jeffries Wyman 
(1875: 14) first pointed out that Ampullaria 
and Paludina shells (now known as Poma- 
cea and Viviparus respectively) much larger 
than those now known, were to be found in 
shell heaps on the St. Johns River. He gives 
measurements which illustrate this clearly. 
Clarence B. Moore substantiated Wyman's 
findings by discovering even larger ex- 
amples. However, he went further, pointing 
out that there has been considerable change 
in shell size, with the oldest and most 
modern being similar in size, while those 
of an intermediate period were larger 
(Moore, 1892: 921-922; 1893: 115). 

It has not been possible to equate pre- 
cisely Moore's old, middle, and late periods 
of shell change with the present chronology 
(Fig. 1), but a rough approximation is pos- 
sible. There is little question that the first 
or oldest period Avith small size shells is the 
equivalent of the preceramic horizon, for 
he clearly states this. The period with the 
largest shells certainly includes in part the 
Orange and Tick Island horizons (see the 
Orange Mound, Moore, 1893: 616). How- 
ever, the other sites listed by Moore as 
having large shells appear to have been oc- 
cupied in part into St. Johns II times. 
Whether these large shells occur in that 
late horizon is not clear. But it does appear 
from the data that at least by some time in 

10 It may be noted that an extinct form of 
mink has been described from Maine shell heaps 
(Prentiss, 1903; Loomis, 1911). 

11 Although the great auk has long been extinct 
in New England, bones of the bird are very com- 
mon in Maine shell heaps (Loomis and Young, 

the St. Johns II horizon, the large size 
shells disappeared and the mollusks re- 
verted to a siz3 comparable to that now 

In association with the large size Vivi- 
parus georgianus of the middle period was 
a new form Viviparus georgianus var. altior 
Pilsbry (1892: 142). Moore (1892: 922) notes 
that the proportions of aperture to height 
of the Viviparus changed, but it is not clear 
whether this new variety represents the 
change or whether it is Avithin the regular 
Viviparus georgianus. 12 

Data on mammals are limited, but deer 
bones from South Indian Field appear to 
represent individuals much larger and more 
massive than the present range of Florida 
deer. The remains of the round-tailed musk- 
rat, Neo fiber alleni True, are also much 
larger than contemporary forms, so much 
so that Houck (MS.) has expressed the pos- 
sibility that they may represent a new 

Climatic change as evidenced by vegetation. 
— The well-known royal palms on the St. 
Johns River reported by Bartram (1940: 
113) are a good example of what may be a 
recent climatic change in the past 200 
years. The modern (circa 1900) natural 
range of the royal palm is twofold. One area 
of scattered occurrences is from Little 
River, Dade County, south and westward 
beyond Royal Palm Park to the Cape 
Sable region. In this spread of distribution 
they are closely associated with tropical 
vegetation. The other region is the southern 
end of Fakahatchee Slough or Strand in 
Collier County. Here the palms flourish in 
great numbers, reaching 80 to 100 feet in 
height, but in a temperate cypress and 
water-oak association. Strangely enough, a 
thatch palm is also found here. Between 
these two areas is the small group of palms 
on Lost mans River. 

The most obvious explanation for the 
presence of the royal palms on the St. 
Johns would be that they made their 
northern extension during a warm period 
and then held on in favorable localities 

12 Comparable changes in proportion have been 
noted in Maine shell heaps and elsewhere in the 
world in other species of shells (Morse, 1882; 



until cold weather finally killed them off, 
moving their range southward. It is pos- 
sible that, once established, they can sur- 
vive a long time in a nontropical environ- 
ment (as in Collier County), and that the 
climatic change that resulted in their ex- 
tinction in north Florida actually was begun 
sometime before Bartram's visit in 1773. 


A series of situations has been presented 
showing changes in the Recent ecology of 
Florida. These were either observations of 
early travelers or are data derived from 
archeological excavation. The changes in- 
cluded differences in sea level, coastal 
ecology, and in flora and fauna. 

At best, much of the value of these data 
is to awaken interest and to stimulate 
further work. Because of the nonscientific 
method of gathering, most of the changes 
described can not be used for critical work. 
However, many of these finds could be du- 
plicated under controlled collecting con- 

The work at South Indian Field can serve 
as a model for further work, especially when 
biological and geological specialists are in- 
cluded in the research program. At this site 
every fragment of animal bone was saved 
and when possible was identified (Masius, 
MS.; Houck, MS.). As a result, strati- 
graphic data, which are tied to the cultural 
sequence can be given for the animal re- 
mains. The detailed survey and testing of 
Rouse (MS.) clearly bring out the unusual 
nature of the site in relation to the physical 

The future of Recent ecological work in 
Florida appears to be promising. The very 
nature of the environment on the border 
between tropical and temperate zones, plus 
a long low coastline must have resulted in 
many changes since the Pleistocene. It is 
probable that this survey has pointed out 
only a small amount of the information to 
be derived in conjunction with archeologi- 
cal work. 

To gain the utmost ecological data from 
archeological work will require real coopera- 
tion. The archeologist whose interest is 
usually only cultural history is loath to 
collect all animal materials. He has found 

from experience that few naturalists are 
equipped to identify or even interested in 
identifying zoological . material, which is 
often in very fragmentary condition. How- 
ever, if given any encouragement he will 
probably be found eager to cooperate in col- 
lecting non-cultural material, for the ar- 
cheological awareness of their value is grow- 


Allison, Ira S. Pluvial lakes and pumice. Sci. 
Monthly 42: 63-65. 1946. 

Baker, Frank C. Pleistocene land and fresh 
water Mollusca as indicators of time and eco- 
logical conditions. In "Early Man," edited 
by G. G. MacCurdy: 67-74. 1937. 

Bartram, William. The travels of William 
Bartram. New York, 1940. 

Bi atchley, W. S. A nature wooing at Ormond by 
the Sea. Indianapolis, 1902. 

Calkins, W. W. Some notes of personal in- 
vestigation among the shell mounds of Flor- 
ida. Proc. Davenport Acad. Nat. Sci. 2: 
225-229. 1878. 

Clench, W. J., and Abbott, R. Tucker. The 
genera Gaza and Livona in the western At- 
lantic. Johnsonia (12). Cambridge, 1943. 

Cressman, L. S. Stratigraphic evidence. Sci. 
Monthly 42: 43-51. 1946. 

Dachnowski-Stokes, Alfred P. Peal profiles 
of the Everglades in Florida The strati- 
graphic features of the "Upper" Everglades and 
correlation with environmental changes. 
Journ. Washington Acad. Sci. 20: 89-106. 

Davis, John H., Jr. The ecology and geologic 
role of mangroves in Florida. Carnegie Inst. 
Washington Publ. 517. 1940. 

. The peat deposits of Florida. Florida 

Geol. Surv. Geol. Bull. 30. Tallahassee, 1946. 

Ecker, A. Zur Kenntniss des Korperbaues 
fruherer Einwohner der Halbinsel Florida. 
Arch, fur Anthrop. 10: 101-114. 1878. 

Eisely, Loren C. Pollen analysis and its bear- 
ing upon American prehistory. Amer. 
Antiq. 5: 115-139. 1939. 

Flint, Richard F. Glacial geology and the 
Pleistocene epoch. New York, 1947. 

Goggin, John M. A preliminary definition of 
archeological areas and periods in Florida. 
Amer. Antiq. 13: 114-127. 1947. 

Glock, W. S. Principles and methods of tree ring 
analysis. Carnegie Inst. Washington Publ. 
486. 1937. 

Hansen, Henry P. Pollen analysis and post- 
glacial climate and chronology. Sci. Monthly 
42:52-62. 1946. 

. Postglacial forest succession, climate, and 

chronology in the Pacific Northwest. Trans. 
Amer. Philos. Soc. 37(1). 1947. 

Haury, E. W. Tree rings, the archeologist' s time 
piece. Amer. Antiq/ 1 : 98-108, 1935. 

July 15, 1948 



Hitchcock, C. H. The Hernandes shell-heap, 
Ormond, Florida. Science 16: 203. 1902. 

Houck, Margaret Van Winkle. Animal re- 
mains in kitchen-middens from South Indian 
Field, Florida. MS., Yale Peabody Mu- 
seum, New Haven, Conn. 

Ihering, Herman von. A origem dos Samba- 
quis. Rev. Inst. Hist, e Geogr. Brazil 8: 
446-457. 1903. 

Johnson, Frederick, et al. The Boylesion 
Street fishweir. Papers Robert S. Peabody 
Foundation for Archeology 2. 1942. 

Johnson, Frederick, and Raup, Hugh M. 
Grassy Island. Archeological and botanical 
investigations of an Indian site in the Taunton 
River, Massachusetts. Papers Robert S. 
Peabody Foundation for Archeology 1(2). 

Koenigswald, Gustav von. Die indianischen 
Muschelberge in Sudbrasilien. Globus 87: 
341-347, 1905. 

Krone, Ricardo. Informacoes ethnographicas 
do valle do Rio Ribeira de Iguape. Bol. Com. 
Geogr. e Geol. Estado de Sao Paulo. 1914. 

Linne, S. Darien in the past. Goteborgs Kungl. 
Vet. Vitter-Samh. Handl. Femte Foljden, 
ser. A, 1(3). 1929. 

Loomis, F. B. New mink from the shell heaps of 
Maine. Amer. Journ. Sci., ser. 4, 31: 227- 
229. 1911. 

Loomis, F. B., and Young, D. B. On the shell 
heaps in Maine. Amer. Journ. Sci., ser. 4, 
34: 17-42. 1912. 

Masius, Vera M. Chronology at South Indian 
Field, Florida. MS., Yale Peabody Mu- 
seum, New Haven Conn. 

Moore, Clarence B. Certain shell heaps of the 
St. John's River, Florida, hitherto unex- 
plored. Amer. Nat. 26: 912-922. 1892. 

. Idem. Amer. Nat. 27: 8-13, 113-117, 

605-624, 708-723. 1893. 

. Idem. Amer. Nat. 28: 15-26. 1894. 

Morse, E. S. Evidences of great antiquity in the 
shell heaps at- Goose Island. Proc. Boston 

Soc. Nat. Hist. 11: 301-302. 1868. 

. Changes in Mya and Lunatia since the 

deposition of the New England shell heaps. 
Proc. Amer. Assoc. Adv. Sci. 30: 345. 1882. 

. Shell-mounds and changes in the shell. 

composing them. Sci. Monthly 21 : 429-440 
Perry, Louise M. Marine shells of the south- 
west coast of Florida. Bull. Amer. Pal. 26 
(95). 1940. 
Pilsbry, Henry A. Preliminary notices of new 
forms of fresh water shells. Nautilus 3: 142- 
143. 1892. 
Prentiss, D. W. Description of an extinct mink 
from the shell heaps of the Maine coast. Proc. 
U. S. Nat. Mus. 26: 887-888. 1903. 
Richards, Horace G. Marine Pleistocene mol- 
lusks as indicators of time and ecological con- 
ditions. In "Early Man," edited by G. G. 
MacCurdy: 75-84. Philadelphia, 1937. 
Romans, Bernard. A concise natural history of 

east and west Florida. New York, 1775. 
Rouse, Irving. An introduction to the archeol- 
ogy of the Melbourne region, Florida. MS., 
Yale Peabody Museum, New Haven, 
Serrano, Antonio. The Sambaquis of the Bra- 
zilian coast. Handbook of South American 
Indians (Bur. Amer. Ethnol. Bull. 143) 1: 
401-407. 1946. 
Simpson, Charles T. Contributions to the Mol- 
lusca of Florida. Proc. Davenport Acad. 
Nat. Sci. 5: 45-72. 1887-89. 
Willey, Gordon R. Excavations in southeast 
Florida. MS., Yale Peabody Museum and 
Bureau of American Ethnology, Smith- 
sonian Institution, New Haven and Wash- 
Wyman, Jeffries. Fresh-water shell mounds of 
the St. John's River, Florida. Peabody 
Acad. Sci. Mem. 4. Salem, 1875. 
Zeuner, Frederick E. Dating the past, an in. 
troduction to geochronology. London, 1946. 

CHEMISTRY. — Purification of uranium oxide. 1 
Bureau of Standards. 

James I. Hoffman, National 

Early in the summer of 1941 Leo Szilard 2 
gave a sample of impure uranyl nitrate to 
the author and requested that the uranium 
be separated from everything else. It was 
his desire to obtain the residue, after the 

1 Received April 28, 1948. The information 
covered in this paper will appear at a later date in 
Division VIII of the Manhattan Project Techni- 
cal Series. 

2 Member of Power Production Subsection of 
the Uranium Committee. See Atomic Energy, by 
Henry D. Smyth, Princeton University Press, 

removal of uranium, for measurements of 
neutron absorption. During the same period 
samples of impure uranium oxide were sent 
to the National Bureau of Standards for 
chemical analysis. For both purposes it was 
desirable to use a solvent that would dis- 
solve the bulk of the uranium without dis- 
solving any of the other substances. The 
solubility of uranyl nitrate in ethyl ether, 
first reported by E. Peligot, 3 suggested its 

3 Ann. Chim. Phys. (3) 5: 5. 1842. 



use for extraction of uranium as the nitrate, 
somewhat as ether is used in steel analysis 
for the extraction of ferric chloride. This 
extraction had been successfully applied in 
1939 for separating uranyl nitrate from 
rhenium and certain rare earths in work on 
the determination of rhenium and molyb- 
denum, 4 but, because of the conflicting 
statements concerning the solubility of the 
rare earth nitrates in ether, 5 it was obvious 
that experimental work was necessary, es- 
pecially since Szilard stated that certain 
members of the rare earth group are strong 
neutron absorbers. 

The work herein described is not intended 
to give the procedure finally used in large- 
scale production of uranium or the methods 
used in the various laboratories of the Man- 
hattan Area. Many improvements in pro- 
cedure have been made since this prelimi- 
nary work was done. 

impurities, but it was evident that extrac- 
tion of uranyl nitrate with ether should be 
a good starting point for the purification of 

To obtain a quick answer to the possi- 
bility of using ether for the purification of 
uranium a nitric-acid solution of the ele- 
ments listed in mixture A 6 in nitric-acid 
solution was evaporated as far as possible 
on the steam bath. The dry residue was ex- 
tracted with 100 ml of ether, the ether was 
washed twice with 5 drops of water, and 
then the uranium in the ether was con- 
verted to oxide. 

Qualitative spectrochemical analysis 
showed that this oxide contained only cal- 
cium, magnesium, and silicon, besides ura- 
nium. These encouraging results prompted 
further experiments, which formed the basis 
for the procedure given in the following 


The bulk of the uranium was extracted 
with ether from 100 g of uranyl nitrate, 
U0 2 (N0 3 )2.6H 2 0, to which had been added 
5 ml of water and "rare earth" nitrates 
(equivalent to 0.0084 g of oxides) contain- 
ing cerium, praseodymium, neodymium, 
erbium, gadolinium, terbium, dysprosium, 
holmium, erbium, ytterbium, lutecium, 
scandium, yttrium, lanthanum, and tho- 
rium. After removal of most of the uranium, 
the combined "rare earths" were deter- 
mined in the residual water solution by pre- 
cipitating them as fluorides, converting the 
fluorides to sulphates, then precipitating as 
oxalates, and igniting the oxalates to "rare 
earth" oxides. These oxides weighed 0.0086 
g. In this particular case the uranium ex- 
tracted by the ether was not examined for 

4 Ether extraction of uranyl nitrate actually 
was first suggested to the author by the work of 
W. F. Hillebrand, U. S. Geol. Surv. Bull. 78: 47. 
1891, and such extractions were in use in our 
laboratory during the 1930's. Although the fact 
that uranyl nitrate could be extracted with ethyl 
ether was known for many years prior to the work 
recorded in this paper (see references in footnotes 
3 and 5), data on the neutron-absorbing elements 
in atomic-energy work were of course not avail- 

5 R. C. Wells, Journ. Washington Acad. Sci. 
20: 146. 1930; F. Soddy and R. Pirret, Phil. 
Mag. (6) 20: 345. 1910; W. F. Hillebrand, U. S. 
Geol. Surv. Bull. 78: 47. 1891; C. W. Davis, 
Amer. Journ. Sci. (5) 11: 20. 1926. 


Fifty grams of U 3 8 was transferred to a 
600-ml beaker, 75 ml of diluted nitric acid 
(1 volume of concentrated nitric acid, sp. gr. 
1.42, diluted with 1 volume of water) was 
added, and the beaker was placed on the 
steambath until all action ceased. The in- 
soluble matter was removed by filtration, 
and the residue was washed four or five 
times with diluted nitric acid (1 volume of 
nitric acid, sp. gr. 1.42, diluted with 20 
volumes of water). The filtrate and wash- 
ings containing the uranyl nitrate were 
evaporated to dryness on the steambath. To 
the cooled and dried residue in the beaker 5 
ml of water and 100 ml of ethyl ether were 
added, and the beaker was swirled until the 
uranyl nitrate was dissolved. The large 
quantities of impurities caused the aqueous 
phase in the bottom of the beaker to have 
the appearance of an emulsion, but this did 
not interfere in separating the layers be- 

6 Mixture A consisted of 2.5 g of uranium metal 
dissolved in an excess of nitric acid. To this was 
added as nitrates 2.5 mg each of copper, anti- 
mony, lead, aluminum, lithium, zirconium, 
indium, gallium, bismuth, potassium, dysprosium, 
cadmium, gadolinium, chromium, magnesium, 
holmium, cobalt, tin, calcium, and 10 mg of a 
mixture of "rare earths" known to contain 
cerium, lanthanum, scandium, praseodymium, 
thorium, and yttrium. 

July 15, 1948 



cause it was possible to pour the ether into a 
separatory funnel without mixing with the 
water layer. 

The beaker was washed three times with 
5-ml portions of ether, which were likewise 
transferred to the separatory funnel. The 

funnel was stoppered and shaken vigorously 
for J to 1 minute. After allowing the liquid 
to stand 3 minutes, a few drops of water 
appeared in the bottom of the separatory 
funnel. This water was drained into the 
beaker that originally contained the ether 

Table 1. — Spectrochemical Analysis 1 of Products Obtained in the Purification of Uranium 
Oxide or Uranium from Ore Concentrates by the Procedure Described Herein 


U:0 8 #155 

U 3 8 #181 

Carnotite ore 2 

Pitchblende ore 2 







ore con- 

U 3 0s 

ore con~ 

U 3 8 




























































































































































































































































































































































































































— ■ 

Note. — The designations VS and S correspond to major constituents (greater than 1 percent); M and W to minor constituents 
(1 to 0.01 percent); and VW, T, and FT to trace constituents (less than 0.01 percent). The absence of a designation indicates that 
a test was not made for that element. 

1 By V. F. Scribner and H. R. Mullin. 

2 The water extracts contained rare earths, but no attempt was made to identify them individually. 



solution. Five ml of water was now added 
to the solution in the separatory funnel, the 
mixture was shaken vigorously, and the 
solution was again allowed to stand until 
two layers formed. The aqueous phase was 
drained into the beaker that originally con- 
tained the ether solution. The washing with 
another 5-ml portion of water was repeated 
once more. 

To convert the purified uranyl nitrate to 
oxide a little water was added to the ether, 
the ether was cautiously evaporated, and 
the residue was ignited to U 3 08 at 1000°C. 
It was afterward found preferable to add 
20 ml of water to the ether solution, shake 
vigorously for 1 minute, and allow the liquid 
to separate into two layers. The water 
layer containing the uranium was drained 
into a suitable dish, and the extraction with 
20-ml portions of water was repeated until 
the ether above the water was colorless. 
Three or four such extractions were suffi- 
cient to remove the uranium. The combined 
water extracts were evaporated to dryness, 
and the uranyl nitrate in the residue was 
ignited to U 3 8 at 1000°C. The ether from 
which the uranyl nitrate had been removed 
was suitable for future extractions. 

The procedure was also applied to the 
extraction of uranium from pitchblende and 
carnotite ore concentrates by digesting the 
ore concentrate with nitric acid, evapo- 
rating to dryness, and extracting the residue 
with ether. The efficacy of this method of 
purification is shown in Table 1. The table 
shows that in many cases impurities that 
were not detected in the original oxide were 
concentrated and detected in the water ex- 
tract. Note especiallv Ce, Co, Cr, Dy, Er, 
Eu, Gd, Ho, La, Lu, Nd, Pr, Sc, Sm, Tb, 
Tm, and Y. 


It was evident that the purification of 
crude uranium oxide and the removal of 
uranium from ore concentrates by conver- 
sion of the uranium to uranyl nitrate and 
extraction with ether had possibilities be- 
cause practically all impurities were re- 
moved in a single operation. Tests by L. F. 
Curtiss 7 indicate that practically none of 
the radium in the original ore was extracted 
by the ether. 

Under stress of wartime conditions not 
all the possible confirmatory tests were 
made, but, as a check on removal of the 
rare earths, the purified oxides obtained 
from no. 155 and no. 181, Table 1, were put 
through a second purification by the same 
procedure. The water extracts in this case 
showed no rare earths. Tests by K. D. 
Fleischer indicated that if 0.5 mg of "rare 
earth" oxides had remained in the purified 
oxide, a positive test would have been ob- 
tained here. These tests showed also that 
by this simple procedure the combined 
"rare earths" in the purified oxide were 
reduced to less than 5 parts per million. 
The spectrochemical tests showed that not 
more than 0.5 part of cadmium or boron 
per million remained in the purified mate- 


A procedure is given for the purification 
of uranium oxide by converting the oxide 
to the nitrate and partitioning the nitrate 
between a large amount of ether and a rela- 
tively small amount of water. In modified 
form, the procedure was found to be appli- 
cable to pitchblende and carnotite ore con- 
centrates, as indicated in Table 1. 

7 Chief of Section on Radioactivity at the Na- 
tional Bureau of Standards. 

BOTANY. — Notes on North American Leguminosae. 1 Frederick J. Hermann. 
U. S. Department of Agriculture. 

In the course of routine identification of 
collections of Leguminosae sent to the U. S. 
Department of Agriculture over a period of 
several years the advisability of proposing 
the following transfers and changes in status 
has become apparent. Included also is a 

1 Received March 25, 1 948. 

diagnosis for an apparently hitherto un- 
described Phaseolus from Mexico. 

Acacia schaffneri (S. Wats.), comb. nov. 

Piihecolobium schaffneri S. Wats., Proc. Amer. 
Acad. 17: 352. 1882; Samanea schaffneri Mac- 
bride, Contr. Gray Herb. 59: 2. 1919; Poponax 

July 15, 1948 Hermann: notes on north American leguminosae 


schaffneri Britton & Rose, N. Amer. Flora 23: 
89. 1928. 

Acacia pinetorum, nom. nov. 

Vachellia peninsularis Small, Man. Southeastern 
Flora 654. 1933; not Acacia peninsularis 
(Britton & Rose) Standi., Field Mus. Publ. Bot. 
11: 158. 1936, based on Senegalia peninsularis 
Britton & Rose, N. Amer. Flora 23: 116. 1928. 

Schrankia angustisiliqua (Britton & Rose), 

comb. nov. 
Leptoglottis angustisiliqua Britton & Rose, N. 
Amer. Flora 23: 143. 1928. 

Schrankia chapmani (Small ex Britton & Rose), 
comb. nov. 

Leptoglottis chapmanii Small ex Britton & Rose, 
N. Amer. Flora 23: 141. 1928. 

Desmanthus pringlei (Britton & Rose), 
comb. nov. 

Acuan pringlei Britton & Rose, N. Amer. Flora 
23: 134. 1928. 

Caesalpinia colimensis, nom. nov. 

Brasilettia glabra Britton & Rose, N. Amer. Flora 
23: 321. 1920, not Caesalpinia glabra (Mill.) 
Merrill, Philipp. Journ. Sci. 5: 54. 1910. 

Caesalpinia pumila (Britton & Rose), 
comb. nov. 

Guaymasia pumila Britton & Rose, N. Amer. 
Flora 23: 322. 1930; Caesalpinia gracilis 
Benth. ex Hemsi., Diag. PI. Nov. 9. 1878, not 
Miq., Fl. Ind. Bat. 1: 110. 1855. 

Dalea tuberculina (Rydb.), comb. nov. 

Parosela tuberculina Rydb., N. Amer. Flora 24: 
89. 1920. 

Petalostemum candidum Michx., var. oligo- 
phyllum (Torr.), comb. nov. 

Petalostemum gracile var. oligophyllum Torr. in 
Emory, Notes Mil. Rec. 139. 1848; P. candidus 
var. occidentalis Gray ex Heller in Britton & 
Kearney, Trans. N. Y. Acad. Sci. 14: 33. 1895; 
Petalostemon oligophyllus "Torr.," Smyth, 
Trans. Kans. Acad. Sci. 15: 61. 1898; P. oc- 
cidental (Gray) Fernald, Rhodora 39: 28. 1937. 

Tephrosia virginiana (L.) Pers., var. leuco- 
sericea (Rydb.), comb. nov. 

Cracca leucosericea Rdyb., N. Amer. Flora 24: 
163. 1923; Tephrosia leucosericea (Rydb.) Cory, 
Rhodora 38: 406. 1936. 

Tephrosia ambigua (M. A. Curtis) Chapm., 
var. intermedia (Small), comb. nov. 

Cracca intermedia Small, Bull. Torr. Bot. Club 21 : 
303. 1894. 

Astragalus tenellus Pursh, var. strigulosus 

(Rydb.), comb. nov. 

Homalobus strigulosus Rydb., Bull. Torr. Bot 
Club 34: 420. 1907; Astragalus tenellus f. 
strigulosus Macbr., Contr. Gray Herb. 65: 34. 

Astralagus michauxii (Kuntze). 
comb. nov. 

Tragacantha michauxii Kuntze, Rev. Gen. 941. 
1891; A. glaber Michx. 1803, not Lam. 1783. 

Astragalus ceramicus Sheld., var. filifolius 
(Gray), comb. nov. 

Astragalus pictus var. filifolius A. Gray, Proc. 
Amer. Acad. 6: 215. 1864; Psoralea longifolia 
Pursh, Fl. Amer. Sept. 741. 1841, not Astra- 
galus longifolius Lam. 1783; Astragalus mito- 
phyllus Kearney, Lean. W. Bot. 4 (8): 216. 

The above combination is necessitated by 
the fact that Gray's Astragalus pictus (Proc. 
Amer. Acad. 6: 214. 1864) is a later homonym 
of the validly published A. pictus Boiss. • & 
Guillardot in Boiss., Diagn. PI. Orient. Ser. 2, 
3 (6): 55. 1859, so that A. ceramicus Sheld. 
must be taken up in its stead for our American 
plant. Although Dr. Kearney's epithet is avail- 
able for Gray's var. filifolius when treated in 
specific rank, the proportion of specimens in- 
termediate between it and A. ceramicus that 
have come to the attention of the writer leads 
to the conclusion tnat varietal status may be 
more appropriate for it. 

Centrosema arenicola (Small), comb. nov. 

Bradburya arenicola Small, Fl. Southeastern U. S. 
651. 1903. 

Desmodium arenicola (Vail), comb. nov. 

Meibomia arenicola Vail, Bull. Torr. Bot. Club 
23: 140. 1896; Hedysarum lineatum Michx., Fl. 
Bor. Amer. 2: 72. 1803, not L. 1759; Des- 
modium lineatum (Michx.) DC, Prodr. 2: 330. 

Although the name Desmodium lineatum 
(Michx.) DC. has been in general use for this 
plant it can be maintained only in spite of 
Article 61. An unfortunate consequence of this 
rule, as it now stands, is that a later homonym 
is rendered permanently hors de combat, even 
though the specific epithet upon transference 
to another genus would not constitute a homo- 
nym in its new context. 



Rhynchosia simplicifolia (Walt.) Wood, 
var. intermedia (T. & G.), comb. nov. 

Rhynchosia iomentosa 13 intermedia T. & G., N. 
Amer. Fl. 1: 285. 1838; R. intermedia (T. & G.) 
Small, Man. Southeastern Flora 715. 1933. 

Phaseolus neglectus, sp. nov. 

Herba volubilis; stipulis lineari-oblongis, 3-5 
nerviis, 5-6 mm longis, rigidis; stipellis lineari- 
oblongis, 2-3 mm longis; foliolis deltoideo- 
acuminatis vulgo plus minusve lobatis; pedun- 
culis 5-11 cm longis: bracteis persistentibus, 
rigidis, 6-9 mm longis, 5-nerviis, subtus pie- 
rum que pilosis: pedicellis tenuibus glabratis; 
bracteolis caducis, uninerviis; calyce campanu- 
lato-cupuliformi; corolla 20 mm longa; vexillo 
obovato, valde emarginato; alis orbiculari- 
ovatis; ovario dense piloso. 

Herbaceous vine; stems slender, sparsely 
puberulous with reflexed hairs to glabrate; 
stipules linear-oblong, 3-5 nerved, 5-6 mm 
long, rigid; petioles puberulous to glabrate, 3-6 
cm long; stipels linear-oblong, rigid, 2-3 mm 
long; leaflets 3, membranaceous, deltoid- 
acuminate, sparsely puberulent above, glabrous 
to sparsely puberulent beneath, 2-6 cm long, 
2-4.5 cm wide, lobed (often only shallowly so 
or even entire), the median 3-lobed, the lateral 
2-lobed, the lobes round-ovate and generally 
shallow; peduncles slender, 5-11 cm long, 11- 
25-flowered; bracts persistent, green, lanceo- 
late-acuminate, firm, 5-nerved, generally more 
or less pilose beneath, sparingly so to glabrate 
above, 6-9 mm long, 1-1.5 mm wide at base; 

pedicels slender, glabrate, 3.5-5 mm long; 
bracteoles caducous, green, narrowly linear- 
lanceolate, one-nerved, glabrous, 2-2.5 mm 
long; calyx campanulate-cupuliform, 2.5-3.5 
(lower lip up to 5) mm long, very sparingly 
ciliolate, the lower lip irregularly pilose, promi- 
nently 3-lobed with median lobe 2.5 mm long, 
acute, upper lip very shallowly 2-lobed; 
corolla 20 mm long, pale salmon to light blue; 
standard obovate, 16 mm long, 12 mm wide, 
deeply emarginate, the upper half reflexed; 
wings orbicular-ovate, abruptly contracted into 
a broad claw, its lower half adnate to the keel; 
keel tubular, with two complete coils; free 
stamen with a reniform enlargement above the 
base; style-beard extending around the first 
coil; stigma lateral; ovary linear, densely 

Nuevo Leon, Mexico in oak woods along 
trail up Sierra de la Cebolla from La Trinidad, 
Municipio de Montemorelos, C. H. Muller 
2881, Aug. 20, 1939 (type— U. S. National 
Arboretum Herbarium). 

Nearest allied to Phaseolus foliaceus Piper, of 
the Sierra Madre. From this it differs in its 
longer, linear-oblong, rather than triangular- 
lanceolate stipules; its 5-nerved, linear-lanceo- 
late bracts which are pilose below; its shallowly 
lobed leaflets; longer peduncles bearing ra- 
cemes with more numerous flowers; campanu- 
late-cupuliform calyx which is almost imper- 
ceptibly ciliolate; corolla 20, rather than 10, 
mm long; obovate, deeply emarginate stand- 
ard; and orbicular-ovate wings. 

of Iowa. 

Two new species of Physarum. 1 G. W. Martin, State University 

The two species of Physarum here noted 
were included in the extensive collections of 
Myxomycetes made by William Bridge 
Cooke on Mount Shasta, Calif., and sub- 
mitted by him for identification. One of 
them proves to be identical with two old 
collections from Mount Rainier, Wash., 
which have been in this laboratory for many 
years awaiting determination. Both appear 
to be clearly distinct from any recognized 
species in this large genus. 

1 Received November 5, 1947. 

Physarum rubronodum, sp. nov. 

Sporangiate, globose to obovate or pulvi- 
nate, sessile or borne on weak, strandlike stalks 
produced as extensions of the hypothallus, 
pinkish brown, or dark when lime is scanty in 
peridium, 1-1.5 mm in diameter, densely 
clustered on a common hypothallus; peridium 
double, the outer layer cartilaginous, calcareous, 
shining, crustose, smooth except for a coarse 
overlying reticulation or, when lime is scanty, 
dark and lacking the reticulation, the inner 
layer membranous, closely applied, colorless, 

July 15, 1948 

martin: two new species of physarum 


iridescent; hypothallus prominent, silvery to 
yellow, venose, the veins often projecting as 
stalk-like extensions on which sporangia are 
borne; capillitium profuse, close-meshed, bear- 
ing large fusiform or irregularly angular scarlet 
or pinkish nodes, most of the junctions lime- 
less; spores nearly black in mass, dark viola- 
ceous brown by transmitted light, slightly 
paler on one side, densely and somewhat ir- 

Fig. 1.- — Physarum rubronodum. Left, group of 
sporangia (Cooke 15671a) on hypothallus, X12; 
right, detail of capillitium with two spores (Cooke 
18126), X520; above, spore of same, X 1,000. 

regularly verrucose, globose, ll-13/i in dia- 
meter, or oval and correspondingly longer and 
narrower. Plasmodium scarlet or orange-red. 

Sporangiis globosis vel obovatis vel pulvi- 
natis, sessilis vel substipitatibus, miniato-brun- 
neis, 1-1.5 mm diam., dense caespitosis sub 
hypothallo commune; peridio duplici, extus 
cartilagineo, calcareo, crustaceo, nitente, laeve 
praeter crasso reticulato; intus membranaceo, 
applicato, hyalino, iridescente; capillitio denso 
e filamentis hyalinis reticulato-anastomosatis, 
nodulis calcareis, magnis, fusiformibus vel 
irregulariter angularibus, coccineis vel miniatis 
multis axillis ecalcareix; sporis globosis vel 
subovoideis, atropurpureo-brunneis crebro 
grosseque tuberculatis 11-13ju diam. Plasmodio 
coccineo vel aurantiaco. 2 

California: Mount Shasta, 8,000 feet, elev., 
July 7, 1941, W. B. Cooke 15671a, type. Same 
locality, June 29, 1947, W. B. Cooke 18126. 

Physarum rubronodum is obviously close to 
P. albescens Macbr. Like that species, it has a 
crustose outer wall suggesting Leocarpus, and 
many of the sporangia are borne on strandlike 
extensions of the hypothallus. It differs in the 

2 I am indebted to Sister Mary Cecelia Bodman 
for assistance with the Latin diagnoses. 

pinkish-brown or dark peridium, in the striking 
scarlet nodes, in the somewhat smaller, warted 
rather than spiny spores, and in the scarlet or 
deep orange-red Plasmodium. 

Physarum auripigmentum, sp. nov. 

Sporangiate, stalked, gregarious; sporangium 
globose, 0.4^0.6 mm in diameter, total height 
0.6-1 mm., clear to opaque yellow (about lemon- 
chrome of Ridgway); peridium membranous, 
closely covered by subcircular limy scales; 
dehiscence somewhat petaloid; columella none; 
stalk short, about half the diameter of the 
sporangium, cylindrical, expanded at the base, 
orange-red, limeless, translucent; hypothallus 
scarcely evident; capillitium dense, delicate, 
persistent, the nodes small, rounded, bright 
yellow, many of the junctions limeless and with 
numerous free, pointed ends; spores dark 
brown in mass, clear violet-brown by trans- 
mitted light, nearly smooth, (8.5-) 9.5-11 
(-12.5)/x in diameter, Plasmodium unknown. 

Sporangiis stipitatibus, gregariis, globosis, 
0.4-0.6 mm diam., totis 0.6-1 mm altis, lucidis 
haud pellucidis luteis; peridiis membranaceis, 
squamis suborbiculatis calcareis applicatis; 
subfloriforme dehiscentibus; columella nulla; 
stipes brevis, attingens dimidium diametrosis 
sporangii, cylindraceus, expansus basi, auran- 
tiacus, ecalcareus, translucidus; hypothallo quo 

Fig. 2. — Physarum auripigmentum (Cooke 
20099). Left below, two sporangia, X60; right, 
detail of capillitium, with two spores, X520; left, 
above, three spores, X 1,000. 

vix sentiri; capillitio denso, delicato, persis- 
tente, nodulis parvulis rotundatis, luteis, mul- 
tis axillis ecalcareis, apicibusque multiplicibus, 
apertis, acutis; sporis atro-brunneis, violaceo- 
brunneis pellucente, sublevibus, 9.5-1 lp diam. 
Plasmodio ignoto. 

California: Mount Shasta, 8,000 feet elev., 
June 23, 1947, W. B. Cooke 20099, type.' 



Washington: Longmire Springs, Mount 
Rainier, August 10-17, 1928, D. B. Creager 
S. U. I. 1722; Paradise Valley, Mount Rainier, 
August, 10-17, 1928, D. B. Creager S. U. I. 

turn Macbr. but differs in the scaly character 
of the peridium, in the shorter and brighter 
stalks without dark basal deposits, and espe- 
cially in the dense capillitium, with its small, 
regular nodes, numerous lime-free junctions 

Physarum auripigmentum suggests P. obla- and conspicuous free ends 

ENTOMOLOGY. — Synoptic revision of the United States scarab beetles of the 
subfamily Dynastinae, No. 5: Keys to tribes and genera. 1 Lawrence W. 
Saylor, California Academy of Sciences. 

This paper completes my studies covering 
a synoptic review of the United States dy- 
nastine scarab beetles, the preceding four 
parts having been published in this same 

A great deal of work, from a taxonomic 
standpoint, remains to be done in the 
American members of this tribe, especially 
in the Neotropical genera. Generic limits of 
such genera as Ligyrus and Stenocrates, as 
well as many others, must be thoroughly 
studied and the relative importance of such 
characters as the front male claws (enlarged 
or not), the dentition of the mandibles, and 
the usual sexual dimorphism must be better 

In the present studies I have had the 
cooperation of many institutions and indi- 
viduals in obtaining material, or submitting 
material for identification: United States 
National Museum, through the courtesy of 
Drs. Wetmore and Chapin; the extensive 
collections of the California Academy of 
Sciences through its director Dr. Miller and 
its entomological curators Drs. Ross and 
Van Dyke; American Museum of Natural 
History through its curator Dr. Cazier; and 
many private individuals, among them 
Drs. Cartwright, Ritcher, Reinhard, and 
Sanderson, as well as Dr. Dampf of Mexico 
City. I have also received material from the 
Paris Museum through Dr. Paulian, and 
from the British Museum through Dr. 
Hinton and Mr. Arrow, to all of whom I am 
indebted for numerous past favqrs. 

Subfamily Dynastinae 

Diagnostic characters. — Tarsal claws always 
equal in size, or at least so on the middle legs 
(one claw of the front pair is frequently en- 

1 Received September 3, 1947. 

larged in males of certain species); mandibles 
entirely corneous, and usually exposed beyond 
the clypeus (from dorsal view); mandibles fre- 
quently large and dentate externally; labrum 
hidden under the clypeus; clypeus more com- 
monly acuminate apically, and dentate or 
edentate; scutellum normal, never greatly en- 
larged; sexes frequently dimorphic, the males 
frequently with tubercles or horns on either 
head or thorax or both, the females in many 
species likewise equipped; coloration usually 
some shade of black or brown, only very rarely 
with any metallic lustre; antennal club always 
relatively small and 3-segmented; ligula en- 
tirely connate with the mentum; abdominal 
spiracles diverging strongly behind; anterior 
coxae transverse, not prominent; stridulating 
organs frequently appear in many species, lo- 
cated on propygidium or inside the elytra; 
fifth ventral sternite and propygidium connate, 
the last spiracle on the suture between them; 
onychium between the tarsal claws commonly 
bisetose, varying to multisetose in certain 


1. Labial palpi inserted behind the mentum; body 

always depressed above; frequently with 
tubercles or horns on head or thorax; mid- 
disc of thorax often longitudinally im- 
pressed; hind tibia digitate or truncate at 
apex but not noticeably widened; first seg- 
ment of hind tarsus with strong spine at 
apex; sexual differences hardly apparent. . . 


Labial palpi inserted at the sides of the ligular 
part of the mentum; body never strongly de- 
pressed, usually evenly convex dorsally: 
head and thorax horned or not; sexual char- 
acters noticeable in last abdominal sternite 
or front tarsal claw in all instances 2 

2. Head and thorax in both sexes entirely un- 

armed, without tubercles or carinae or horns, 
and never depressed or foveate; claw with 
the onychium always bisetose (never more 

July 15, 1948 saylor: synoptic revision of subfamily dynastinae 


than two cilia); stridulating organs absent; 
male front claws in many species either en- 
larged or larger than in female; antenna 
often larger in male than female; tarsi 

cylindrical and usually elongate, never tri- 
angular; prosternal spine prominent behind 

the fore coxae Cyclocephalini 

Head and thorax (either or both) armed with 

Fig. 1. — a, Ligyrus relictus Say: Hind tibia of male; 6, Euetheola subglabra Schaeffer: Hind tibia 
of male; c, Aphonus densicauda Casey: Hind tibia of male; d, Cheiroplatys clunalis LeConte: Hind 
tibia of male; e, Xyloryctes jamaicensis (Drury): Worn hind tibia of male; /, same as e, fresh tibia; g, 
Strategics antaeus (Drury): Hind tibia of male; h, Strategics mormon Burmeister: Hind tibia of male. 



horns or tubercles (on fore margin of thorax 
if nowhere else) in both sexes, ard frequently 
foveate (or with strong head carina and 
large, well-rounded mandibles as in Aphon- 
ides); claw with onychium bisetose to multi- 
setose; stridulating organs variable; other 

characters variable 3 

3. Onvchium between claws bisetose to pluri- 
setose; tarsal segments, especially basally, 
frequently triangular in shape; fore tibia and 
tarsus of same length in both sexes; stridulat- 
ing organs frequently present; dorsal surface 
always unicolorous; never spotted or with 

dense hair Oryctini 

Onychium between claws always with three or 
more setae apically ; tarsal segments elongate 
smooth, never triangular, the basal segment 
usually with a strong apical spine; front 
tibia and tarsus a little longer in the male 
than in the female (very distinct, in neo- 
tropical species, but must be carefully com- 
pared in United States species) ; stridulating 
organs absent; dorsal surface either clothed 
with a fine velvety short hair, or spotted and 
speckled, onl., rarely entirely unicolorous in 
some females Dynastimi 

The genera of the tribes Phileurini and 
Dynastini were covered in Part 4 of this 
series, and the genera of the tribe Cycloce- 
phalini in Part 1. 2 The tribe Oryctini was 
discussed in Part 2 3 and in Part 3 4 and was 
completed in Part 4 5 ; these eight United 
States genera of Oryctini may be separated 
as follows: 


1. Apex of hind tibia uneven, with sharp angula- 

tions or teeth (see Fig. 1, g, /); head fre- 
quently with horn (male) or large tubercle 
(female) (this character will place those 
specimens with worn tibia as in some Xy- 

loryctes, as in Fig. 1, e 2 

Apex of hind tibia moderately to strongly ex- 
panded, with very fine serrations (not sharp 
teeth) or entirely smooth (see Fig. 1, a, d); 
head never with horns, at most with a strong 
to weak transverse carina or a small tubercle 

2. Clypeal apex acute and unidentate, mandibles 

large and exposed, and evenly rounded, never 
toothed; thorax never foveate or tubercu- 
late; base of clypeus with a strong transverse 

carina (Texas and Arizona) 

Aphonides Rivers 

Clypeus acute, bidentate or bluntly rounded 
(if acute, then the mandibles always dis- 
tinctly emarginate or toothed externally, 

2 Journ. Washington Acad. Sci. 35 (12) : 378- 
386. 1945. 

3 Ibid. 36 (1): 16-22. 1946. 

4 Ibid. 36 (2): 41-46. 1946. 

5 Ibid. 38 (5): 176-183. 1948. 

and thorax always foveate and tuberculate) 

3. Mandibles usually hidden beneath the clypeus, 

or only edges exposed, and always unarmed 
externally; usually with 5 to 9 small teeth 

on outer apical margin of hind tibia 

Xyloryctes Hope 

Mandibles always large, always well exposed, 
always armed externally with teeth or else 
right-angled in outline (cessus); apical mar- 
gin of hind tibia usually with 1 or 2 sharp 

angulations, or 3 or 4 teeth 

Strategus Hope 

4. Clypeal apex acute, unidentate; clypeal base 

with an acute transverse carina; mandibles 
large and tridentate; thorax with fovea at 
midapex and a small tubercle in front of the 

fovea; color usually rufous 

Oxygrylius Casey 

Clypeal apex distinctly bidentate, or evenly 
truncate, or very bluntly rounded, never 
with a single sharp point (or if somewhat 
pointed, then the mandibles small and nearly 
hidden under the clypeus) ; color variable . . 5 

5. Clypeus with a distinct carina just before (i.e., 

practically on) the apex, this carina entire, 
or wide bidentate, or tridentate; front male 
tibia frequently edentate, the tibia unusually 
wide in both sexes and the emarginations 
between the external teeth very shallowly 
indicated, especially the two apical teeth. .6 
Clypeus carinate or not; if carinate the carina 
located quite a distance before the apex; 
front tibia always strongly tridentate .... 7 

6. Preapical clypeal carina entire, or bidentate; 

thorax and head tuberculate or not (South- 
western United States). Cheiroplatys Hope 
Preapical clypeal carina strongly to weakly 
tridentate; thorax not tuberculate, head 

rarely so (Eastern United States) 

Aphonus LeConte 

7. Mandibles definitely tridentate externally, the 

basal tooth often worn and barely visible; 
thorax not more than two-thirds the length 
of the elytra; color rufous to rufopiceous or 

rufocastaneous Ligyrus Burmeister 

Mandibles bidentate externally (do not count 
lacinia tooth as mandibular!); thorax about 
three-fourths length of elytra; color always 
black Euetheola Bates 

Euetheola subglabra Schaeffer 
Since Part 2 of this series was published I have 
received a male specimen from Tepic, Mexico, 
from Dr. Chapin, of the U. S. National Museum; 
this has very worn mandibles and clypeus, but 
the genitalia are typical, and the thoracic punc- 
turation is very minute and sparse but still visible 
on the center disc in a good light. I have also just 
recently mounted an additional pair (male and 
female) from "Compostella, Tepic, Nayarit, col- 
lected VI-24-40 at light by Morgan M. and L. W, 
Saylor"; these were mixed in with some un- 

July 15, 1948 gahan: south American chalcidoidea described by ashmead 


mounted Dyscinetus material, which they greatly 
resemble superficially. The female has not been 
described before and I am designating this speci- 
men as the Allotype: Very similar to male ex- 
cept that it is a little larger (16 mm), the minute 
thoracic punctures are entirely absent over most 
of the center-disc, the apical half of the pygidium 
is smooth and very sparsely punctate and the 

apical half of the sixth sternite is entirely smooth 
and impunctate. It is interesting that nearly the 
entire abdomen (except for a single transverse 
setigerous row on each sternite near sides) and 
the metasternum (except at sides) are highly 
polished and entirely impunctate in both sexes. 
Both specimens remain in the Saylor Collection 
at the California Academy of Sciences. 

ENTOMOLOGY.— The Herbert H. Smith collection of South American Chalci- 
doidea described by W. H. Ashmead. 1 A. B. Gahan, U. S. Bureau of Ento- 
mology and Plant Quarantine. (Communicated by E. A. Chapin.) 

In his Classification of the chalcid flies or 
the super family Chalcidoidea (Mem. Carne- 
gie Mus. 1 (No. 4, pt. 2): 394-551. 1904) 
W. H. Ashmead published results of his 
study of the Herbert H. Smith collection of 
South American chalcidoids, including a 
bibliographic catalogue of all the known 
South American species. Many new species 
were described, most of them based upon 
material in the Smith collection, but several 
of the included new species were based upon 
material in the United States National 
Museum, collected by Albert Koebele. In 
the bibliographic catalogue, citations were 
given for a number of species previously 
described by Ashmead from material in the 
Smith collection, these descriptions appear- 
ing in Insect Life 3: 456. 1891; Proc. Ent. 
Soc. Washington 3: 233. 1895, and 4: 11- 
14. 1896; and Proc. U. S. Nat. Mus. 22: 
368-375. 1900. Also included were Ash- 
mead's identifications of a number of spe- 
cies described by Walker, Westwood, 
Fabricius, Perty, and other authors. 

Holotypes of 8 species and duplicate par- 
atypes of 36 additional species, as well as 
duplicates of some of the determined spe- 
cies, were retained by Ashmead from the 
Smith collection. The remainder of the col- 
lection was returned to the Carnegie Mu- 
seum, which had acquired it by purchase 
from the collector. There the collection re- 
mained practically undisturbed until 1934, 
when through the kind cooperation of the 
late Dr. Hugo Kahl, then curator of the 
section of entomology at Carnegie Museum, 
and with the consent of the Museum direc- 
tor, I was permitted to bring the whole 

1 Received March 7, 1948. 

collection to the United States National 
Museum for study. Though it still remained 
the property of Carnegie Museum, it has 
reposed in the National Museum since that 
date and while being used for reference has 
been kept, until very recently, in the orig- 
inal boxes and arrangement in which Ash- 
mead left it. 

Early in 1947, through collaboration with 
Dr. George E. Wallace, successor to Dr. 
Kahl at the Carnegie Museum, and with 
the approval of the authorities of both the 
Carnegie Museum and the National Mu- 
seum, arrangements were completed where- 
by, in exchange for a representative collec- 
tion of named North American chalcidoids, 
the entire Smith collection covered by Ash- 
mead's paper, with the exceptions of one- 
half the duplicates in any series of para- 
types or of determined specimens of any 
species, became the property of the National 
Museum. With consummation of the ex- 
change the Carnegie Museum received a 
named collection of North American ma- 
terial comprising 412 species represented by 
1,222 specimens and also a return of 22 
paratypes representing 16 species and 84 
specimens representing 19 determined spe- 
cies from the South American material. 
The National Collection acquires named 
representatives of 248 species made up of 
192 holotypes, 79 paratypes (8 of which 
represent species the holotypes of which 
were already in the National Museum), and 
156 determined specimens representing 48 
species. Also included are 122 unidentified 
specimens, making a total of 549 specimens. 

Unfortunately, the unique types of Par- 
encyrtus brasiliensis Ashmead, Pelorotelus 
coeruleus Ashmead, Horismenus corumbae 



Ashmead, Euplectrus brasiliensis Ashmead, 
and Stenomesius dimidiatus Ashmead have 
become dislodged from the mounting points 
and lost. Since the first two species are gen- 
otypes, their loss is doubly unfortunate. 
Also missing from the collection is the speci- 
men (or specimens) of Idoleupehnus annuli- 
cornis Ashmead recorded from Santarem, 
Brazil. This species was originally described 
(Proc. Ent. Soc. Washington A: 13. 1896) 
from the Island of St. Vincent, West Indies, 
and the holotype is presumably in the Brit- 
ish Museum. 

The material from the Smith collection 
acquired by the National Museum has now 
been removed from the original boxes and 
placed in standard trays and drawers. While 
making the transfer I made the following 
notes regarding certain species: 

Hemitorymus thoracicus Ashmead. — In my 
opinion this is not a monodontomerine. The 
type appears to be a typical Torymus and the 
genus Hemitorymus, therefore, a synonym. 
New synonymy. 

Pie siostig modes brasiliensis Ashmead. — The 
unique type is a male, not a female as stated in 
the description. 

Spilochalcis tarsalis Ashmead. — The type is 
from Corumba instead of Chapada. 

Spilochalcis persimilis Ashmead. — Type lo- 
cality Santarem instead of Chapada. 

Spilochalcis bimaculata Ashmead. — De- 
scribed only in the key to species. The type, a 
female, is from Trinidad. 

Spilochalcis nigropleuralis Ashmead. — The 
holotype female is from Trinidad. The only 
male associated with this female in the collec- 
tion is from Corumba (not Chapada), Brazil, 
and does not entirely fit the description given 
of the male. It does not bear a name label and 
may or may not be the allotype. 

Spilochalcis mayri Ashmead. — The type is 
from Corumba, not Chapada. 

Spilochalcis urichi Ashmead. — No male is 
described, but included among the 12 speci- 
mens in the type series is one male. This is 
colored almost exactly like the females and not 
easily distinguished. 

Spilochalcis lineocoxalis Ashmead. — The 
unique holotype is a male, not a female. 

Spilochalcis chapadensis Ashmead. — In the 
key to species (p. 427) and on the name label, 

the name of this species is chapadensis, but 
heading the formal description (p. 443) it is 
given as chapadae. That the latter spelling is a 
lapsus is obvious, since the name chapadae had 
already been used for another species of Spilo- 
chalcis (p. 432). The name chapadensis should 
be retained for the species on p. 443. 

Spilochalcis persimilis Ashmead. — The 
unique holotype is a male instead of a female. 

Octosmicra nigromaculata Ashmead. — The 
locality is Corumba instead of Chapada. 

Heptasmicra lineaticoxa Ashmead. — One 
paratype is from Maruru, the type and 2 para- 
types from Santarem. 

Metadontia similis Ashmead. — The unique 
holotype is a female instead of a male. 

Pentasmicra brasiliensis Ashmead. — De- 
scribed only in the key to genera. See article by 
Hugo Kahl (Ann. Carnegie Mus. 13: 265. 

Axima brevicornis Ashmead. — The male al- 
lotype and 4 male paratypes of this species do 
not appear to be the same species as the female 
holotype and the female paratype. Instead, 
these males all seem to me to be A. koebelci 

Aximopsis morio Ashmead. — The type local- 
ity for this genotype species, not recorded by 
Ashmead, is Santerem, Brazil. 

Isomodes brasiliensis Ashmead. — Only 4 of 
the 5 specimens recorded by Ashmead were 
found, the specimen from Santarem being 

Prodecatoma bruneiventris Ashmead. — The 
type series consists of 3 females and 1 male. In 
my opinion the 3 females represent 3 different 
species. The male allotype is labeled "Prode- 
catoma flavescens Ashmead. cf Type," but since 
no male was described for flavescens and since 
this male specimen agrees with the description 
of the male as given for bruneiventris, it is ob- 
vious that the labeling is a lapsus. 

Neorilyaflavipes Ashmead. — The type series, 
of which only 9 specimens can be located, is a 
composite made up of apparently 3 different 

Rileya orbitalis Ashmead. — The unique holo- 
type female is from Chapada instead of San- 
tarem. A male in the National Museum col- 
lection from Chapada is labeled "Rileya orbi- 
talis cf, Type No. 8080." Ashmead may have 
originally associated this with the female holo- 
type, but since no male is mentioned in the 

July 15, 1948 deignan: continental races of pycnonotus dispar 


description it should not be considered a para- 
type. This male is apparently not the same 
species as the holotype but appears to be identi- 
cal with Riley a insularis Ashmead. 

Lelaps affinis Ashmead. — A paratype, 
U.S.N.M. no. 8087, is not the same species as 
the holotype but belongs instead to Lelaps 
ferruginea Ashmead. 

Chalcedectes annulipes Ashmead. — Holotype 
from Corumba in May. Paratype from Cha- 
pada already in National Museum collection. 

Macreupelmus brasiliensis Ashmead. — Only 
3 of the 4 specimens mentioned by Ashmead 
located. U.S.N.M. type, no. 8091. 

Anastatus auriceps Ashmead. — Holotype 
from Corumba. Paratype from Chapada miss- 
ing from pin. 

Trichencyrtus chapadae Ashmead (pp. 291, 
392) (= robustus Ashmead, p. 495). See com- 
ments by Gahan and Peck, Journ. Washington 
Acad. Sci. 36: 317. 1946; also Gomes, Bol. Soc. 
Brazil. Agronomia 5: 287. 1942. 

Metopon brasiliensis Ashmead. — The holo- 
type female is from Corumba. The 2 paratype 
females are from Santarem and appear to be not 
the same species as the type. The allotype male, 
also from Santarem, is missing from the pin. 

Horismenus brasiliensis Ashmead. — A speci- 
men in the National Museum from Chapada 
labeled "Type No. 8094" is not the same spe- 
cies as the holotype, which is from Rio de 

Janeiro. This specimen from Chapada is not 
mentioned in the description and hence is 
neither type nor paratype. 

Ametallon chapadae Ashmead. — The holo- 
type from the Smith collection is in poor con- 
dition. The paratype which was in the National 
Museum and recorded in the type book under 
no. 8097 is missing from the pin. 

Trichoporus melleus Ashmead. — The holo- 
type acquired with the Smith collection is a 
female from Santarem. The alleged male de- 
scribed by Ashmead and bearing U.S.N.M. 
type, no. 8098 is a female from Chapada and 
is not the same species as the holotype. 

Trichoporus viridicyaneus Ashmead. — All 20 
specimens of the type series located, 3 of them 
in the National Museum under type no. 8099. 
These 3 are the same species as the holotype, 
but at least 5 of the paratypes in the Smith 
collection appear to be a different species. 

Alophus flavus Ashmead (p. 353, 365) 
( = brasiliensis Ashmead, p. 520). See comments 
by Gahan and Peck, Journ. Washington Acad. 
Sci. 36: 314. 1946. Holotype and allotype from 
Chapada and one paratype from Rio de Janeiro 
in Smith collection. Holotype and paratype are 
alike, but the allotype male is a different spe- 
cies. Also two alleged paratypes in the National 
Museum, type no. 8100, are not the same as 
the holotype and represent two different 

ORNITHOLOGY. — Continental races of the bulbul Pycnonotus dispar (Horsfield). 
H. G. Deignan, U. S. National Museum. 

Geographical variation in continental 
populations of Pycnonotus dispar ("Oto- 
compsa flaviventris" of the older authors) 
appears in a tendency to darker coloration 
of the upperparts, shortening of the crest, 
and reduced wing length from north and 
west to east and south. The effects of wear 
on the plumage are so marked that only 
fresh-plumaged specimens are suitable for 
taxonomic study, and the change from one 
form to another is so gradual that the dis- 
tinctions between juxtaposed races can be 
observed only in series. The difficulty of 
establishing satisfactory limits of range for 

1 Published by permission of the Secretary of 
the Smithsonian Institution. Received January 
16, 1948. 

the several subspecies is complicated by the 
appearance of winter-wandering examples 
of one within the breeding territory of an- 
other. Two hundred and one specimens from 
the Asiatic mainland, brought together in 
Washington with the kind cooperation of 
the authorities of the Academy of Natural 
Sciences of Philadelphia (A.N.S.P.) and of 
the Chicago Natural History Museum 
(C.N.H.M.), enable me to accept the fol- 
lowing forms: 

1. Pycnonotus dispar flaviventris (Tickell) 

Vanga Flaviventris Tickell, Journ. Asiat. Soc. 
Bengal 2: 573. Nov. 1833 (Dampara, "in Dhol- 
bhum" = Singhbhum District, Chota Nagpur 
Division, Province of Bihar and Orissa, India). 



Diagnosis. — Head and neck blue-black; up- 
perparts golden olive-green, somewhat brighter 
posteriorly; exposed portions of remiges and 
rectrices blackish brown, edged with golden 
olive-green; underparts rich yellow, somewhat 
suffused with olive on the breast and along the 

Wing length.— 86-91 mm. (5 males), 85-89 
mm. (4 females), 85-96 mm. (4 unsexed). 

Specimens examined. — "India": No definite 
locality (3 unsexed). Bengal: Darjiling Dis- 
trict: Sivok (1 male, 2 females). Assam: No 
definite locality (1 unsexed); Cachar District: 
Rupachena (1 female); Lakhimpur District: 
Margherita (1 male, 1 female). Burma: Myit- 
kyina District: N'Pon Village (1 male); Pakok- 
ku District: Dudaw Taung (2 males). 

2. Pycnonotus dispar vantynei, 
n. subsp. 

Type.— C.N.H.M. no. 79247, adult male, 
collected at Phongtho (lat. 22°32' N., long. 
103°20 / E.), Laokay or Laichau Province, 
northwestern Tongking, on February 24, 1929, 
by J. Van Tyne (original number 396). 

Diagnosis. — As large as P. d. flaviventris, but 
easily separable by having the upperparts a 
darker and duller olive-green, almost without 
golden suffusion. 

Wing length.— 85-91 mm (20 males), 85-87 
mm (4 females). 

Specimens examined. — Annam: No definite 
locality (1 male); Thanhhoa Province: Hoixuan 
(1 female), Lunglunh (1 male). Tongking: 
Laokay or Laichau Province: Phongtho (2 
males); Laokay Province: Ba Nam Nhung 
(1 male); Laichau Province: Muong Mo (1 
male, 2 females), Muong Moun (1 male), 
Paham (1 male), Laichau (1 male), 27 km 
WSW. of Laichau (2 males). Laos: 5 e Terri- 
toire Militaire: Bountai (3 males), Muong Yo 
(1 male, 1 female); Pakse Province: Pakse (1 
male, winter). Burma: Kengtung State: Loi 
Mwe (3 males, 1 female), Mong Len (1 male); 
Siam: North: Chiang Saen Kao (2 females), 
Chiang Rai (1 male), Muang Fang (1 male), 
Wiang Pa Pao (1 female), Doi Suthep (2 males, 
winter); East: Ban Chanuman (1 male, winter). 

Remarks. — Only one of these examples shows 
the red-throated phase of plumage (discussed 
below under P. d. johnsoni). 

This race is named in honor of Josselyn Van 
Tyne, collector of the type specimen. 

3. Pycnonotus dispar xanthops, n. subsp* 

Type.— U.S.N. M. no. 332362, adult male, 
collected at "Khan River' 1 = Ban Mae (lat. 
18°40' N., long, 98°50' E.), Chiang Mai Prov- 
ince, northwestern Siam, on February 8, 1932, 
by Hugh M. Smith (original number 5321). 

Diagnosis. — Smaller than either of the two 
precedent races and with the color of the upper- 
parts darker, more greenish, less golden, than 
in flaviventris, but much lighter and more 
golden than in vantynei. 

Wing length.— 83-87 mm. (16 males), 83-85 
mm. (4 females). 

Specimens examined. — Siam: Northwest: Doi 
Chiang Dao (1 male), Muang Pai (1 male), 
30 km N. of Chiang Mai (1 male), Doi Suthep 
(5 males, 3 females), vicinity of Chiang Mai 
(3 males, 1 female), Ban Mae (2 males), Doi 
Khun Tan (6 males). Burma: Amherst District 
Myawadi (1 male), Kawkareik (1 female). 

Remarks. — This race presumably has a much 
more extensive range than my material shows; 
it should be the form of all southern Burma. 

The few skins available from southwestern 
Siam are smaller than xanthops; their plumage 
is so badly worn that, for the present, they must 
be left unnamed. 

4. Pycnonotus dispar auratus, n. subsp. 

Type.— A.N.S.P. no. 115110, adult male, 
collected at Amphoe Wat Pa = Muang Lorn 
Sak, on the borders of central and eastern Siam 
at lat. 16°45' N., long 101°10' E., on October 
26, 1934, by collectors for R. Meyer de Schau- 
ensee (original number 1705). 

Diagnosis. — Smaller than either vantynei or 
flaviventris and with the color of the upperparts 
lighter, less greenish, more golden, than in the 
former, but much darker and less golden than 
in the latter. 

From P. d. xanthops, with which it agrees in 
size, auratus is separable by having the upper 
parts slightly darker and less golden. 

Wing length.— 85-87 mm (4 males), 82-84 
mm (5 females). 

Specimens examined. — Laos : Vientiane Prov- 
ince: Vientiane (1 female). Siam: Central or 
East: Muang Lorn Sak (4 males, 5 females). 
Annam: Phanrang Province: Krongpha (1 
female, winter). 

Remarks. — P. d. auratus and P. d. xanthops 
are equivalent "links" in the cline between 
flaviventris and vantynei. The range of auratus 

July 15, 1948 deignan: continental races of pycnonotus dispar 


may be expected to cover the more eastern 
provinces of northern Siam, the northern half 
of the eastern Siamese plateau, and portions of 
Haut- and Moyen-Laos between the territories 
of vantynei and johnsoni. 

One of my specimens shows the red-throated 
phase of plumage; another has an irregular 
spot of red and yellow feathers at the left side 
of the chin. 

5. Pycnonotus dispar johnsoni (Gyldenstolpe) 
Rubigula johnsoni Gyldenstolpe, Kungl. Svenska 
Vet.-Akad. Handl. 50 (8): p. 25, col. pi. 1. July 
19, 1913 (near Sathani Chakkrarat, Nakhon 
Ratchasima Province, eastern Siam at lat. 
15W N., long. 102°25' E.). 

Diagnosis. — Near to P. d. auratus, but smaller 
and with the upperparts slightly darker and 
less suffused with golden. 

Wing length.— 79-85 mm (33 males), 76-83 
mm (34 females), 81-83 mm (3 unsexed). 

Specimens examined. — Siam: East: Sathani 
Pak Chong (7 males, 9 females), Sathani Hin 
Lap (3 males), Sathani Lat Bua Khao (5 males, 
3 females), Ban Tha Chang (1 unsexed), Ban 
Khanong Phra (2 females), Ban Pang Sok (1 
male), Nakhon Nayok (1 unsexed), Ban Chanu- 
man (4 females), Ban Khulu (1 unsexed), 
Ban Khemmarat (1 female); Southeast: 
Khao Saming (2 males, 1 female), Ban Bang 
Phra (1 male, 1 female), Ban Laem Ngop (1 
male), Khao Sa Bap (3 males, 1 female), Khao 
Soi Dao (1 male), Muang Rayong (2 females), 
Siracha (1 male), Ban Nong Kho (3 males), 
Ban Nong Yang (4 males); Laos: Pakse Prov- 
ince: Pakse (2 males, 4 females), Paksong (1 
female, 1 unsexed), Bassac (1 female). Cam- 
bodia: Kampot Province: Ok Yam (1 female); 
Siemreap Province: Banteai Srei (1 female); 
Kompong Cham Province: Kompong Cham (1 
female). Annam: Phanrang Province: Krong- 
pha (1 female, 1 unsexed), between Bellevue 
and Daban (1 male, 3 females), Fimnon (1 fe- 
male), Ban Methuot (1 male). 

Remarks. — A color phase in which the lower 
throat is clothed in shining red feathers is fre- 
quent in johnsoni; in the population of the 
southwestern corner of the eastern Siamese 
plateau (topotypical), all the specimens seen 
show the character, which appears with dimin- 
ishing frequency as one moves away from this 
center and is then often indicated merely by a 
few red feathers among the black. Since red- 
throated individuals appear in any predomi- 

nantly black-throated population of southeast- 
ern Indo-China (and even in populations other- 
wise representative of the races auratus and 
vantynei), it is not possible to restrict the use of 
Gyldenstolpe's name to red-throated birds 
without seeming to have two subspecies resi- 
dent at the same localities, and there is, of 
course, no question here of two sympatric 
species. A case precisely similar to this is pre- 
sented by the bulbul Pycnonotus atriceps (Tem- 
minck), in which the several color phases ap- 
pear with varying frequ